diff --git a/branch_cleanup_guard.py b/branch_cleanup_guard.py index 64db84a..e91ea02 100644 --- a/branch_cleanup_guard.py +++ b/branch_cleanup_guard.py @@ -525,6 +525,90 @@ def assess_ownership_record_activity(record: dict[str, Any]) -> dict[str, Any]: } +# Reviewer-lease reclaim is only reachable from a non-live (expired/stale) lease. +_RECLAIMABLE_REVIEWER_STATUSES = _EXPIRED_STATUSES | _STALE_STATUSES + + +def is_active_ownership_status(status: str | None) -> bool: + """True when *status* denotes live/active ownership of a branch (#855). + + Used to decide whether a *competing* active claimant still uses a branch + when weighing an expired reviewer lease for reclaim. Expired, stale, + released, and terminal statuses are not active. + """ + return _norm_str(status).lower() in _ACTIVE_OWNERSHIP_STATUSES + + +def assess_expired_reviewer_lease_reclaim( + *, + role: str, + status: str, + pr_merged: bool | None, + owner_pid_alive: bool | None, + competing_active_claimant: bool | None, +) -> dict[str, Any]: + """Decide, explicitly and fail-closed, whether an expired reviewer lease + may stop protecting an already-merged branch (#855 AC4). + + An expired reviewer lease should not protect a merged branch forever once + its work is done and no live claimant remains. Reclaim is permitted only + when **every** condition below is provably satisfied; any unknown + (``None``) or contrary value keeps the lease protective: + + - the lease is a ``reviewer`` lease (author/merger/controller/reconciler + leases are out of scope and always keep protecting); + - its status is expired or stale (never an active/live lease); + - the PR is proven merged (``pr_merged is True``); + - the lease owner process is proven dead (``owner_pid_alive is False``); + - no competing active claimant uses the branch + (``competing_active_claimant is False``). + + Returns a decision dict with ``reclaim_allowed`` and, when refused, the + fail-closed ``reasons``. The reasons never contain secrets — only the + role, the status, and which condition was unproven. + """ + reasons: list[str] = [] + normalized_role = _norm_str(role).lower() + normalized_status = _norm_str(status).lower() + + if normalized_role != "reviewer": + reasons.append( + f"lease role '{normalized_role or 'unknown'}' is not a reviewer " + "lease; expired-reviewer reclaim does not apply" + ) + if normalized_status not in _RECLAIMABLE_REVIEWER_STATUSES: + reasons.append( + f"lease status '{normalized_status or 'unknown'}' is not expired " + "or stale; only a non-live reviewer lease may be reclaimed" + ) + if pr_merged is not True: + reasons.append( + "PR merged state is not proven true; reclaim requires an " + "already-merged PR (fail closed)" + ) + if owner_pid_alive is not False: + reasons.append( + "lease owner process liveness is not proven dead; a live owner " + "still protects the branch (fail closed)" + ) + if competing_active_claimant is not False: + reasons.append( + "a competing active claimant may still use the branch; reclaim " + "requires no other active ownership (fail closed)" + ) + + allowed = not reasons + return { + "reclaim_allowed": allowed, + "role": normalized_role, + "status": normalized_status, + "decision": ( + "reclaim_expired_reviewer_lease" if allowed else "keep_protecting" + ), + "reasons": [] if allowed else reasons, + } + + def assess_active_branch_ownership( *, remote: str, diff --git a/canonical_thread_handoff.py b/canonical_thread_handoff.py index aa0a85b..6d262c2 100644 --- a/canonical_thread_handoff.py +++ b/canonical_thread_handoff.py @@ -46,6 +46,17 @@ _FIELD_RE = re.compile( ) +def is_known_cth_type(value: str | None) -> bool: + """True when *value* is a declared member of the :data:`CTH_TYPES` contract. + + ``CTH_TYPES`` is the single authority for what a CTH type may be. The + heading a comment carries is free text, so a *read* path that turns a parsed + type into something durable — a serialized field, a routing decision — must + check membership here rather than trust the parse or keep a list of its own. + """ + return (value or "").strip() in CTH_TYPES + + def format_cth_body( *, cth_type: str, @@ -60,7 +71,7 @@ def format_cth_body( ) -> str: """Render a canonical CTH comment body.""" normalized_type = (cth_type or "").strip() - if normalized_type not in CTH_TYPES: + if not is_known_cth_type(normalized_type): raise ValueError( f"unknown CTH type '{cth_type}'; expected one of {sorted(CTH_TYPES)}" ) @@ -101,6 +112,12 @@ def parse_cth_comment(body: str) -> dict[str, Any] | None: fields[key] = match.group(2).strip() return { "cth_type": cth_type, + # The heading capture is unconstrained free text, so the parse states + # whether it satisfies the CTH_TYPES contract instead of leaving every + # reader to decide (or forget). Parsing stays total — an unknown type is + # still parsed and reported, never raised on — but a reader that turns + # the type into a durable value can now tell the two apart. + "cth_type_known": is_known_cth_type(cth_type), "fields": fields, "raw_body": text, } @@ -119,7 +136,7 @@ def assess_cth_comment(body: str) -> dict[str, Any]: } cth_type = parsed.get("cth_type") or "" - if cth_type not in CTH_TYPES: + if not is_known_cth_type(cth_type): reasons.append( f"unknown CTH type '{cth_type}'; expected one of {sorted(CTH_TYPES)}" ) diff --git a/control_plane_db.py b/control_plane_db.py index 75af7c9..4616cb3 100644 --- a/control_plane_db.py +++ b/control_plane_db.py @@ -599,6 +599,35 @@ class ControlPlaneDB: (_ts(), session_id), ) + def list_sessions( + self, + *, + statuses: Sequence[str] | None = None, + limit: int = 500, + ) -> list[dict[str, Any]]: + """List session rows for restart / impact analysis (#658). + + Read-only. Sessions are the process-level unit an MCP restart + disrupts, so the restart coordinator inventories them to compute blast + radius. Optional ``statuses`` filter (e.g. ``('active',)``) narrows to + live rows. Never returns secrets — only operational metadata. + """ + clauses: list[str] = [] + params: list[Any] = [] + if statuses: + placeholders = ", ".join("?" for _ in statuses) + clauses.append(f"status IN ({placeholders})") + params.extend(statuses) + where = ("WHERE " + " AND ".join(clauses)) if clauses else "" + sql = ( + f"SELECT * FROM sessions {where} " + "ORDER BY last_heartbeat_at DESC LIMIT ?" + ) + params.append(max(1, int(limit))) + with self._tx(immediate=False) as conn: + rows = conn.execute(sql, params).fetchall() + return [dict(r) for r in rows] + # ── work items ──────────────────────────────────────────────────────── def upsert_work_item( diff --git a/dirty_orphan_worktree_recovery.py b/dirty_orphan_worktree_recovery.py new file mode 100644 index 0000000..17db425 --- /dev/null +++ b/dirty_orphan_worktree_recovery.py @@ -0,0 +1,1080 @@ +"""Dirty orphaned author-issue worktree recovery (#860). + +Self-hosting deadlock class observed against #850 / PR #853 and #855: + +* same-claimant durable issue lock (``jcwalker3`` / ``prgs-author``) +* registered dirty worktree under ``branches/`` +* lock missing PID / session PID / expiry / heartbeat (malformed) +* existing recovery (#753/#768/#772) and renewal require a *clean* worktree + and/or a determinable owner PID +* no sanctioned dirty-preserving rebind + sync to a newer remote PR head + +This module is the pure evidence assessor and crash-safe recovery orchestrator +for that one class. It is an **explicit** recovery operation — it does not +silently widen ``gitea_lock_issue``. + +Safety model +------------ +Eligibility requires *all* of: + +* same claimant identity and profile as the durable lock +* exact issue / repository / branch / registered source worktree agreement +* registered worktree under the canonical branches root (resolved-path ancestry) +* no active original process (when PID is present) +* no active competing author session or workflow lease +* sufficient corroborating evidence when PID fields are absent (caller pins) +* explicit caller-provided local head, remote/PR head, and dirty fingerprints +* no foreign, duplicate, or ambiguous ownership + +A PID-less lock is **never** considered live merely because expiration fields +are absent (see also ``issue_lock_store.assess_lock_freshness``). + +Dirty preservation +------------------ +The source worktree is frozen: never cleaned, reset, overwritten, or deleted. +Recovery prefers a separately prepared recovery worktree checked out at the +pinned remote PR head. Dirty bytes are re-applied with path-level conflict +detection when upstream also changed a dirty path. + +Crash safety +------------ +A durable journal is written *before* filesystem or ownership mutation. +Retries resume or fail closed without stealing ownership, duplicating +worktrees, or losing dirty bytes. +""" + +from __future__ import annotations + +import hashlib +import json +import os +import shutil +import stat +import subprocess +from dataclasses import dataclass +from typing import Any, Mapping, Sequence + +from issue_lock_store import is_process_alive +from reviewer_worktree import parse_dirty_tracked_files + +# Outcomes +ELIGIBLE = "ELIGIBLE" +REFUSED = "REFUSED" +NO_CANDIDATE = "NO_CANDIDATE" +RECOVERY_COMPLETED = "RECOVERY_COMPLETED" +RECOVERY_RESUMED = "RECOVERY_RESUMED" +CONFLICTS_PRESENT = "CONFLICTS_PRESENT" + +# Journal phases (ordered) +PHASE_ELIGIBILITY = "1_eligibility_proven" +PHASE_JOURNAL_PERSISTED = "2_journal_persisted" +PHASE_RECOVERY_WORKTREE = "3_recovery_worktree_prepared" +PHASE_DIRTY_APPLIED = "4_dirty_applied" +PHASE_BINDING = "5_session_bound" +PHASE_COMPLETE = "6_complete" + +JOURNAL_DIR_NAME = "dirty-orphan-recovery-journals" +REQUIRED_LOCK_FIELDS = ("issue_number", "branch_name", "worktree_path") + +# Marker directory written into recovery worktrees for governed conflicts. +CONFLICT_STATE_DIR = ".gitea-recovery" +CONFLICT_STATE_FILE = "conflicts.json" + + +def _text(value: Any) -> str: + return str(value or "").strip() + + +def _same_realpath(left: str | None, right: str | None) -> bool: + if not left or not right: + return False + try: + return os.path.realpath(left) == os.path.realpath(right) + except OSError: + return left == right + + +def sha256_bytes(data: bytes) -> str: + return hashlib.sha256(data).hexdigest() + + +def sha256_file(path: str) -> str: + with open(path, "rb") as fh: + return sha256_bytes(fh.read()) + + +def get_journal_dir(override: str | None = None) -> str: + if override: + path = override + elif os.environ.get("GITEA_DIRTY_ORPHAN_RECOVERY_JOURNAL_DIR"): + path = os.environ["GITEA_DIRTY_ORPHAN_RECOVERY_JOURNAL_DIR"] + else: + path = os.path.join( + os.path.expanduser("~/.cache/gitea-tools"), JOURNAL_DIR_NAME + ) + os.makedirs(path, mode=0o700, exist_ok=True) + return path + + +def _journal_path(idempotency_key: str, journal_dir: str | None = None) -> str: + safe = "".join( + c if c.isalnum() or c in ("-", "_", ".") else "_" + for c in idempotency_key + ) + return os.path.join(get_journal_dir(journal_dir), f"{safe}.json") + + +def load_journal( + idempotency_key: str, journal_dir: str | None = None +) -> dict[str, Any] | None: + path = _journal_path(idempotency_key, journal_dir=journal_dir) + if not os.path.isfile(path): + return None + if os.path.islink(path): + raise ValueError(f"refusing journal path that is a symlink: {path}") + with open(path, "r", encoding="utf-8") as fh: + data = json.load(fh) + if not isinstance(data, dict): + raise ValueError("corrupt recovery journal: not an object") + return data + + +def save_journal( + journal: Mapping[str, Any], journal_dir: str | None = None +) -> str: + key = _text(journal.get("idempotency_key")) + if not key: + raise ValueError("journal requires idempotency_key") + path = _journal_path(key, journal_dir=journal_dir) + if os.path.islink(path): + raise ValueError(f"refusing to write journal through symlink: {path}") + tmp = f"{path}.tmp.{os.getpid()}" + with open(tmp, "w", encoding="utf-8") as fh: + json.dump(dict(journal), fh, indent=2, sort_keys=True) + fh.flush() + os.fsync(fh.fileno()) + os.replace(tmp, path) + return path + + +def derive_idempotency_key( + *, + remote: str, + org: str, + repo: str, + issue_number: int, + source_worktree: str, + expected_local_head: str, + expected_remote_head: str, +) -> str: + raw = "|".join( + [ + "dirty-orphan-recovery", + _text(remote), + _text(org), + _text(repo), + f"issue-{int(issue_number)}", + os.path.realpath(_text(source_worktree)), + _text(expected_local_head)[:40], + _text(expected_remote_head)[:40], + ] + ) + return hashlib.sha256(raw.encode("utf-8")).hexdigest()[:32] + + +def _lock_claimant(lock: Mapping[str, Any]) -> dict[str, str]: + claimant = lock.get("claimant") + if not isinstance(claimant, Mapping): + lease = lock.get("work_lease") + claimant = lease.get("claimant") if isinstance(lease, Mapping) else None + if not isinstance(claimant, Mapping): + return {} + return { + "username": _text(claimant.get("username")), + "profile": _text(claimant.get("profile")), + } + + +def _recorded_pid(lock: Mapping[str, Any]) -> Any: + pid = lock.get("session_pid") + if pid is None: + pid = lock.get("pid") + return pid + + +def _malformed_pid_fields(lock: Mapping[str, Any]) -> list[str]: + """Return reasons the lock's PID fields are unusable (not automatically live).""" + missing: list[str] = [] + pid = _recorded_pid(lock) + if pid is None or _text(pid) == "": + missing.append("session_pid/pid") + return missing + try: + if int(pid) <= 0: + missing.append("session_pid/pid") + except (TypeError, ValueError): + missing.append("session_pid/pid") + return missing + + +def is_path_under_canonical_branches( + path: str, + *, + canonical_repo_root: str, + branches_dirname: str = "branches", +) -> tuple[bool, list[str]]: + """Resolved-path ancestry under ``{canonical_repo_root}/branches``. + + Rejects string-substring tricks, traversal, and symlink escapes outside + the canonical branches root. + """ + reasons: list[str] = [] + if not path or not canonical_repo_root: + return False, ["path and canonical_repo_root are required"] + try: + repo_root = os.path.realpath(canonical_repo_root) + branches_root = os.path.realpath(os.path.join(repo_root, branches_dirname)) + # realpath on a non-existent path still normalizes; prefer it so + # assessment can run without the directory existing yet. + target = os.path.realpath(path) + except OSError as exc: + return False, [f"path resolution failed: {exc}"] + + if not target.startswith(branches_root + os.sep) and target != branches_root: + reasons.append( + f"path '{path}' is not under canonical branches root '{branches_root}'" + ) + return False, reasons + try: + rel = os.path.relpath(target, branches_root) + except ValueError: + return False, ["path not relative to branches root"] + if rel.startswith(".."): + return False, ["path escapes branches root via relative traversal"] + return True, [] + + +def _result( + outcome: str, + *, + eligible: bool, + reasons: list[str], + evidence: dict[str, Any], +) -> dict[str, Any]: + return { + "outcome": outcome, + "eligible": eligible, + "recovery_eligible": eligible, + "reasons": list(reasons), + "evidence": evidence, + } + + +def assess_dirty_orphan_recovery( + existing_lock: Mapping[str, Any] | None, + *, + issue_number: int, + branch_name: str, + source_worktree_path: str, + remote: str, + org: str, + repo: str, + identity: str, + profile: str, + expected_local_head: str, + expected_remote_head: str, + expected_dirty_fingerprints: Mapping[str, str], + current_branch: str | None, + porcelain_status: str, + observed_local_head: str | None, + observed_remote_head: str | None, + observed_dirty_fingerprints: Mapping[str, str] | None, + competing_live_locks: Sequence[Mapping[str, Any]] | None = None, + competing_live_sessions: Sequence[Mapping[str, Any]] | None = None, + workflow_lease_active: bool | None = None, + workflow_lease_expired: bool | None = None, + canonical_repo_root: str, + worktree_registered: bool, + current_pid: int | None = None, + owner_process_alive_override: bool | None = None, +) -> dict[str, Any]: + """Assess whether dirty-orphan recovery is eligible. Pure; no I/O mutation.""" + evidence: dict[str, Any] = { + "issue_number": issue_number, + "branch_name": branch_name, + "source_worktree_path": source_worktree_path, + "remote": remote, + "org": org, + "repo": repo, + "identity": identity, + "profile": profile, + "expected_local_head": _text(expected_local_head), + "expected_remote_head": _text(expected_remote_head), + "expected_dirty_paths": sorted(expected_dirty_fingerprints or {}), + } + reasons: list[str] = [] + + if not existing_lock: + return _result( + NO_CANDIDATE, + eligible=False, + reasons=["no existing durable lock for this issue"], + evidence=evidence, + ) + + lock = dict(existing_lock) + if lock.get("issue_number") != issue_number: + return _result( + NO_CANDIDATE, + eligible=False, + reasons=[ + f"existing lock targets issue #{lock.get('issue_number')}, " + f"not #{issue_number}" + ], + evidence=evidence, + ) + + for field in REQUIRED_LOCK_FIELDS: + if not _text(lock.get(field)): + reasons.append(f"durable lock missing required field '{field}'") + + # Repository / branch / worktree agreement + for field, expected in (("remote", remote), ("org", org), ("repo", repo)): + actual = _text(lock.get(field)) + if actual and actual != _text(expected): + reasons.append( + f"lock {field} '{actual}' does not match requested '{_text(expected)}'" + ) + elif not actual: + # Some legacy locks omit remote/org/repo; require explicit pin match + # via caller still supplying them and branch/worktree agreement. + evidence[f"lock_{field}_absent"] = True + + locked_branch = _text(lock.get("branch_name")) + if locked_branch != _text(branch_name): + reasons.append( + f"lock branch '{locked_branch}' does not match requested " + f"'{_text(branch_name)}'" + ) + evidence["locked_branch"] = locked_branch + + locked_wt = _text(lock.get("worktree_path")) + if not _same_realpath(locked_wt, source_worktree_path): + reasons.append( + f"lock worktree '{locked_wt}' does not match declared source " + f"'{_text(source_worktree_path)}'" + ) + evidence["locked_worktree_path"] = locked_wt + + under, under_reasons = is_path_under_canonical_branches( + source_worktree_path, canonical_repo_root=canonical_repo_root + ) + if not under: + reasons.extend(under_reasons) + if not worktree_registered: + reasons.append("source worktree is not registered in git worktree list") + + # Claimant identity + claimant = _lock_claimant(lock) + evidence["lock_claimant"] = claimant + if claimant.get("username") != _text(identity): + reasons.append( + f"foreign claimant identity '{claimant.get('username')}' " + f"(caller '{_text(identity)}')" + ) + if claimant.get("profile") != _text(profile): + reasons.append( + f"foreign claimant profile '{claimant.get('profile')}' " + f"(caller '{_text(profile)}')" + ) + + # PID / liveness + pid_missing = _malformed_pid_fields(lock) + recorded_pid = _recorded_pid(lock) + evidence["recorded_pid"] = recorded_pid + evidence["pid_fields_missing"] = pid_missing + if pid_missing: + evidence["pid_less_malformed"] = True + # PID-less is never live by missing expiry alone. Eligibility continues + # only with full corroborating pins (already required below). + else: + alive = ( + owner_process_alive_override + if owner_process_alive_override is not None + else is_process_alive(int(recorded_pid)) + ) + evidence["owner_process_alive"] = alive + if alive: + reasons.append( + f"original owner process pid {recorded_pid} is still alive; " + "recovery refused" + ) + + # Competing ownership + for entry in competing_live_locks or (): + if not isinstance(entry, Mapping): + continue + if entry.get("issue_number") == issue_number: + reasons.append( + "competing live lock observed for the same issue; recovery refused" + ) + for entry in competing_live_sessions or (): + if not isinstance(entry, Mapping): + continue + sess_issue = entry.get("issue_number") + sess_identity = _text(entry.get("identity") or entry.get("username")) + if sess_issue == issue_number and sess_identity and sess_identity != _text(identity): + reasons.append( + f"competing live author session by '{sess_identity}' on issue " + f"#{issue_number}" + ) + elif sess_issue == issue_number and entry.get("active"): + # same claimant active elsewhere still blocks ambiguous ownership + if entry.get("session_pid") not in (None, current_pid, os.getpid()): + reasons.append( + "ambiguous competing same-issue author session still active" + ) + + if workflow_lease_active is True and workflow_lease_expired is not True: + reasons.append( + "active competing workflow lease still live; recovery refused" + ) + evidence["workflow_lease_active"] = workflow_lease_active + evidence["workflow_lease_expired"] = workflow_lease_expired + + # Dirty requirement (this recovery class is *for* dirty trees) + dirty_files = parse_dirty_tracked_files(porcelain_status) + if not dirty_files and not expected_dirty_fingerprints: + reasons.append( + "worktree is clean and no dirty fingerprints were provided; " + "use clean-worktree recovery (#753/#772) instead" + ) + evidence["observed_dirty_files"] = dirty_files + + # Explicit pins + if not _text(expected_local_head) or len(_text(expected_local_head)) < 40: + reasons.append("expected_local_head pin missing or not a full SHA") + if not _text(expected_remote_head) or len(_text(expected_remote_head)) < 40: + reasons.append("expected_remote_head pin missing or not a full SHA") + if not expected_dirty_fingerprints: + reasons.append("expected_dirty_fingerprints pin is required") + + if _text(observed_local_head) and _text(observed_local_head) != _text( + expected_local_head + ): + reasons.append( + f"local head mismatch: observed {_text(observed_local_head)} != " + f"pinned {_text(expected_local_head)}" + ) + if _text(observed_remote_head) and _text(observed_remote_head) != _text( + expected_remote_head + ): + reasons.append( + f"remote/PR head mismatch: observed {_text(observed_remote_head)} != " + f"pinned {_text(expected_remote_head)}" + ) + if _text(expected_local_head) == _text(expected_remote_head): + # Divergence is the motivating case; equal heads are allowed only when + # dirty files still need rebinding, so do not refuse equality. + evidence["heads_equal"] = True + else: + evidence["heads_diverged"] = True + + if current_branch and _text(current_branch) != locked_branch: + reasons.append( + f"source worktree is on branch '{_text(current_branch)}', not " + f"locked branch '{locked_branch}'" + ) + + # Fingerprint verification + observed_fps = dict(observed_dirty_fingerprints or {}) + for path, expected_fp in (expected_dirty_fingerprints or {}).items(): + rel = _text(path) + if not rel or rel.startswith("/") or ".." in rel.split("/"): + reasons.append(f"unsafe dirty path pin refused: {path!r}") + continue + obs = _text(observed_fps.get(rel)) + if not obs: + reasons.append(f"missing observed fingerprint for dirty path '{rel}'") + elif obs != _text(expected_fp): + reasons.append( + f"dirty fingerprint mismatch for '{rel}': " + f"observed {obs} != pinned {_text(expected_fp)}" + ) + + # PID-less corroboration: all explicit pins must already have passed. + if pid_missing and reasons: + reasons.append( + "PID-less malformed lock additionally requires full pin corroboration; " + "one or more corroborating checks failed" + ) + + if reasons: + return _result(REFUSED, eligible=False, reasons=reasons, evidence=evidence) + + evidence["eligibility"] = ELIGIBLE + return _result(ELIGIBLE, eligible=True, reasons=[], evidence=evidence) + + +def detect_path_conflicts( + *, + dirty_paths: Sequence[str], + local_head_contents: Mapping[str, bytes | None], + remote_head_contents: Mapping[str, bytes | None], + dirty_contents: Mapping[str, bytes], +) -> list[dict[str, Any]]: + """Path-level conflicts: upstream and dirty patch both changed the path.""" + conflicts: list[dict[str, Any]] = [] + for path in dirty_paths: + local_b = local_head_contents.get(path) + remote_b = remote_head_contents.get(path) + dirty_b = dirty_contents.get(path) + if dirty_b is None: + continue + # Upstream changed relative to the local head version of the path. + upstream_changed = (local_b or b"") != (remote_b or b"") + dirty_differs_from_remote = dirty_b != (remote_b or b"") + if upstream_changed and dirty_differs_from_remote: + conflicts.append( + { + "path": path, + "local_head_sha256": sha256_bytes(local_b) if local_b is not None else None, + "remote_head_sha256": sha256_bytes(remote_b) if remote_b is not None else None, + "dirty_sha256": sha256_bytes(dirty_b), + "reason": ( + "upstream and preserved dirty patch both changed this path" + ), + } + ) + return conflicts + + +def _safe_open_lockfile(path: str): + """Open a lock file refusing symlinks (O_NOFOLLOW when available).""" + if os.path.islink(path): + raise ValueError(f"refusing lock file that is a symlink: {path}") + flags = os.O_RDWR | os.O_CREAT + if hasattr(os, "O_NOFOLLOW"): + flags |= os.O_NOFOLLOW + fd = os.open(path, flags, 0o600) + try: + st = os.fstat(fd) + if stat.S_ISLNK(st.st_mode): + os.close(fd) + raise ValueError(f"refusing lock file that is a symlink: {path}") + except Exception: + try: + os.close(fd) + except OSError: + pass + raise + return fd + + +def build_recovery_lock_record( + *, + existing_lock: Mapping[str, Any], + issue_number: int, + branch_name: str, + recovery_worktree_path: str, + remote: str, + org: str, + repo: str, + identity: str, + profile: str, + expected_remote_head: str, + source_worktree_path: str, + conflicts: Sequence[Mapping[str, Any]], + session_pid: int, +) -> dict[str, Any]: + """Construct a new durable lock bound to the recovery worktree + live PID.""" + from datetime import datetime, timedelta, timezone + + now = datetime.now(timezone.utc) + expires = now + timedelta(hours=4) + record = { + "remote": remote, + "org": org, + "repo": repo, + "issue_number": issue_number, + "branch_name": branch_name, + "worktree_path": recovery_worktree_path, + "pid": session_pid, + "session_pid": session_pid, + "claimant": {"username": identity, "profile": profile}, + "work_lease": { + "operation_type": "author_issue_work", + "issue_number": issue_number, + "pr_number": existing_lock.get("work_lease", {}).get("pr_number") + if isinstance(existing_lock.get("work_lease"), Mapping) + else existing_lock.get("pr_number"), + "branch": branch_name, + "worktree_path": recovery_worktree_path, + "claimant": {"username": identity, "profile": profile}, + "created_at": now.strftime("%Y-%m-%dT%H:%M:%SZ"), + "expires_at": expires.strftime("%Y-%m-%dT%H:%M:%SZ"), + "last_heartbeat_at": now.strftime("%Y-%m-%dT%H:%M:%SZ"), + }, + "dirty_orphan_recovery": { + "recovered": True, + "source_worktree_path": source_worktree_path, + "recovery_worktree_path": recovery_worktree_path, + "accepted_head": expected_remote_head, + "conflicts": list(conflicts), + "source_frozen": True, + }, + "lock_provenance": { + "source": "gitea_recover_dirty_orphaned_issue_worktree", + "written_by_tool": "gitea_recover_dirty_orphaned_issue_worktree", + "written_at": now.strftime("%Y-%m-%dT%H:%M:%SZ"), + "claimant": {"username": identity, "profile": profile}, + }, + } + # Preserve prior assignment/lease ids when present (non-authoritative). + for key in ("assignment_id", "lease_id", "owner_session", "expected_base_sha"): + if key in existing_lock: + record[key] = existing_lock[key] + return record + + +def write_conflict_state( + recovery_worktree: str, conflicts: Sequence[Mapping[str, Any]] +) -> str: + state_dir = os.path.join(recovery_worktree, CONFLICT_STATE_DIR) + os.makedirs(state_dir, mode=0o700, exist_ok=True) + path = os.path.join(state_dir, CONFLICT_STATE_FILE) + payload = { + "conflicts": list(conflicts), + "resolution": "author_edit_required" if conflicts else "none", + } + with open(path, "w", encoding="utf-8") as fh: + json.dump(payload, fh, indent=2, sort_keys=True) + return path + + +def apply_dirty_bytes( + *, + recovery_worktree: str, + dirty_contents: Mapping[str, bytes], + conflict_paths: set[str], +) -> list[str]: + """Write non-conflicting dirty bytes into the recovery worktree. + + Conflicting paths are written as ``*.recovered-dirty`` siblings so the + original dirty bytes remain recoverable without overwriting upstream. + """ + written: list[str] = [] + for rel, data in dirty_contents.items(): + rel = _text(rel) + if not rel or rel.startswith("/") or ".." in rel.split("/"): + raise ValueError(f"unsafe relative path: {rel!r}") + dest = os.path.join(recovery_worktree, rel) + parent = os.path.dirname(dest) + if parent: + os.makedirs(parent, exist_ok=True) + if rel in conflict_paths: + sidecar = dest + ".recovered-dirty" + with open(sidecar, "wb") as fh: + fh.write(data) + written.append(rel + ".recovered-dirty") + else: + with open(dest, "wb") as fh: + fh.write(data) + written.append(rel) + return written + + +@dataclass +class GitOps: + """Injectable git operations for tests.""" + + def run(self, args: list[str], *, cwd: str) -> subprocess.CompletedProcess[str]: + return subprocess.run( + args, + cwd=cwd, + capture_output=True, + text=True, + check=False, + ) + + +def prepare_recovery_worktree( + *, + canonical_repo_root: str, + recovery_worktree_path: str, + branch_name: str, + remote_head: str, + git_ops: GitOps | None = None, +) -> dict[str, Any]: + """Create or resume a recovery worktree at the pinned remote head. + + Leaves the source worktree untouched. Uses ``git worktree add`` only when + the recovery path does not already exist (idempotent resume). + """ + git = git_ops or GitOps() + under, reasons = is_path_under_canonical_branches( + recovery_worktree_path, canonical_repo_root=canonical_repo_root + ) + if not under: + return {"success": False, "reasons": reasons, "created": False} + + if os.path.isdir(recovery_worktree_path): + # Resume: verify HEAD matches pin. + probe = git.run( + ["git", "rev-parse", "HEAD"], cwd=recovery_worktree_path + ) + head = (probe.stdout or "").strip() + if probe.returncode != 0 or head != remote_head: + return { + "success": False, + "created": False, + "resumed": True, + "head": head, + "reasons": [ + f"resume recovery worktree HEAD '{head}' does not match expected remote HEAD '{remote_head}'" + ], + } + return { + "success": True, + "created": False, + "resumed": True, + "head": head, + "reasons": [], + } + + # Create detached-at-head worktree. Do not run git checkout -B because + # the source worktree holds the branch name (Git exit 128). + add = git.run( + [ + "git", + "worktree", + "add", + "--detach", + recovery_worktree_path, + remote_head, + ], + cwd=canonical_repo_root, + ) + if add.returncode != 0: + return { + "success": False, + "created": False, + "reasons": [ + f"git worktree add failed: {(add.stderr or add.stdout or '').strip()}" + ], + } + return { + "success": True, + "created": True, + "resumed": False, + "head": remote_head, + "reasons": [], + } + + +def run_dirty_orphan_recovery( + *, + assessment: Mapping[str, Any], + existing_lock: Mapping[str, Any], + issue_number: int, + branch_name: str, + source_worktree_path: str, + recovery_worktree_path: str, + remote: str, + org: str, + repo: str, + identity: str, + profile: str, + expected_local_head: str, + expected_remote_head: str, + expected_dirty_fingerprints: Mapping[str, str], + dirty_contents: Mapping[str, bytes], + local_head_contents: Mapping[str, bytes | None], + remote_head_contents: Mapping[str, bytes | None], + canonical_repo_root: str, + bind_lock: bool, + lock_writer: Any | None = None, + git_ops: GitOps | None = None, + journal_dir: str | None = None, + session_pid: int | None = None, + interrupt_after_phase: str | None = None, +) -> dict[str, Any]: + """Execute recovery with crash-journal phases. Idempotent on retry.""" + if not assessment.get("eligible"): + return { + "success": False, + "performed": False, + "outcome": assessment.get("outcome") or REFUSED, + "reasons": list(assessment.get("reasons") or ["not eligible"]), + "evidence": dict(assessment.get("evidence") or {}), + } + + # Verify dirty bytes match pins before any mutation. + for path, expected_fp in expected_dirty_fingerprints.items(): + data = dirty_contents.get(path) + if data is None: + return { + "success": False, + "performed": False, + "outcome": REFUSED, + "reasons": [f"dirty content missing for pinned path '{path}'"], + "evidence": {}, + } + if sha256_bytes(data) != _text(expected_fp): + return { + "success": False, + "performed": False, + "outcome": REFUSED, + "reasons": [f"dirty content fingerprint drift for '{path}'"], + "evidence": {}, + } + + idem = derive_idempotency_key( + remote=remote, + org=org, + repo=repo, + issue_number=issue_number, + source_worktree=source_worktree_path, + expected_local_head=expected_local_head, + expected_remote_head=expected_remote_head, + ) + journal = load_journal(idem, journal_dir=journal_dir) or { + "idempotency_key": idem, + "issue_number": issue_number, + "branch_name": branch_name, + "source_worktree_path": os.path.realpath(source_worktree_path), + "recovery_worktree_path": recovery_worktree_path, + "expected_local_head": expected_local_head, + "expected_remote_head": expected_remote_head, + "expected_dirty_fingerprints": dict(expected_dirty_fingerprints), + "phase": None, + "artifacts_created": { + "journal": False, + "recovery_worktree": False, + "dirty_applied": False, + "binding": False, + }, + "conflicts": [], + "complete": False, + } + + if journal.get("complete"): + return { + "success": True, + "performed": False, + "outcome": RECOVERY_RESUMED, + "reasons": ["recovery already complete; idempotent no-op"], + "evidence": { + "journal": journal, + "recovery_worktree_path": journal.get("recovery_worktree_path"), + }, + "journal": journal, + } + + # Phase 1: eligibility already proven by caller assessment. + journal["phase"] = PHASE_ELIGIBILITY + if interrupt_after_phase == PHASE_ELIGIBILITY: + return { + "success": False, + "performed": False, + "outcome": "INTERRUPTED", + "reasons": ["interrupted after eligibility (test harness)"], + "journal": journal, + "evidence": {"phase": PHASE_ELIGIBILITY}, + } + + # Phase 2: persist journal BEFORE filesystem/ownership mutation. + journal["phase"] = PHASE_JOURNAL_PERSISTED + journal["artifacts_created"]["journal"] = True + save_journal(journal, journal_dir=journal_dir) + if interrupt_after_phase == PHASE_JOURNAL_PERSISTED: + return { + "success": False, + "performed": True, + "outcome": "INTERRUPTED", + "reasons": ["interrupted after journal persistence (test harness)"], + "journal": journal, + "evidence": {"phase": PHASE_JOURNAL_PERSISTED}, + } + + # Phase 3: recovery worktree at remote head. + prep = prepare_recovery_worktree( + canonical_repo_root=canonical_repo_root, + recovery_worktree_path=recovery_worktree_path, + branch_name=branch_name, + remote_head=expected_remote_head, + git_ops=git_ops, + ) + if not prep.get("success"): + journal["phase"] = PHASE_RECOVERY_WORKTREE + journal["last_error"] = prep.get("reasons") + save_journal(journal, journal_dir=journal_dir) + return { + "success": False, + "performed": True, + "outcome": REFUSED, + "reasons": list(prep.get("reasons") or ["recovery worktree failed"]), + "journal": journal, + "evidence": prep, + } + if prep.get("created"): + journal["artifacts_created"]["recovery_worktree"] = True + journal["phase"] = PHASE_RECOVERY_WORKTREE + save_journal(journal, journal_dir=journal_dir) + if interrupt_after_phase == PHASE_RECOVERY_WORKTREE: + return { + "success": False, + "performed": True, + "outcome": "INTERRUPTED", + "reasons": ["interrupted after recovery worktree creation (test harness)"], + "journal": journal, + "evidence": prep, + } + + # Phase 4: conflict detection + dirty apply (source frozen). + conflicts = detect_path_conflicts( + dirty_paths=list(expected_dirty_fingerprints.keys()), + local_head_contents=local_head_contents, + remote_head_contents=remote_head_contents, + dirty_contents=dirty_contents, + ) + conflict_paths = {c["path"] for c in conflicts} + written = apply_dirty_bytes( + recovery_worktree=recovery_worktree_path, + dirty_contents=dirty_contents, + conflict_paths=conflict_paths, + ) + conflict_state_path = write_conflict_state(recovery_worktree_path, conflicts) + journal["conflicts"] = list(conflicts) + journal["written_paths"] = written + journal["conflict_state_path"] = conflict_state_path + journal["artifacts_created"]["dirty_applied"] = True + journal["phase"] = PHASE_DIRTY_APPLIED + # Prove source worktree still exists and was not deleted. + journal["source_still_present"] = os.path.isdir(source_worktree_path) + save_journal(journal, journal_dir=journal_dir) + + # #860 F4: Do NOT finalize session binding if conflicts remain. + if conflicts: + return { + "success": False, + "performed": True, + "outcome": CONFLICTS_PRESENT, + "reasons": [ + "conflicts present: manual resolution required before session binding (fail closed)" + ], + "conflicts": list(conflicts), + "recovery_worktree_path": recovery_worktree_path, + "source_worktree_path": source_worktree_path, + "source_frozen": True, + "journal": journal, + "evidence": { + "written_paths": written, + "conflict_state_path": conflict_state_path, + "accepted_head": expected_remote_head, + }, + } + + # Phase 5: bind session (optional for pure assessor tests). + pid = session_pid if session_pid is not None else os.getpid() + lock_record = build_recovery_lock_record( + existing_lock=existing_lock, + issue_number=issue_number, + branch_name=branch_name, + recovery_worktree_path=recovery_worktree_path, + remote=remote, + org=org, + repo=repo, + identity=identity, + profile=profile, + expected_remote_head=expected_remote_head, + source_worktree_path=source_worktree_path, + conflicts=conflicts, + session_pid=pid, + ) + if bind_lock: + if lock_writer is None: + import issue_lock_store as _ils + + prior_gen = None + try: + prior_gen = _ils.lock_generation(dict(existing_lock)) + except Exception: + prior_gen = None + _ils.bind_session_lock( + lock_record, + expected_generation=prior_gen, + recovery_sanctioned=True, + ) + else: + lock_writer(lock_record) + journal["artifacts_created"]["binding"] = True + journal["phase"] = PHASE_BINDING + save_journal(journal, journal_dir=journal_dir) + if interrupt_after_phase == PHASE_BINDING: + return { + "success": False, + "performed": True, + "outcome": "INTERRUPTED", + "reasons": ["interrupted after binding (test harness)"], + "journal": journal, + "evidence": {"lock_record": lock_record}, + } + + journal["phase"] = PHASE_COMPLETE + journal["complete"] = True + save_journal(journal, journal_dir=journal_dir) + + return { + "success": True, + "performed": True, + "outcome": RECOVERY_COMPLETED, + "reasons": [], + "conflicts": [], + "recovery_worktree_path": recovery_worktree_path, + "source_worktree_path": source_worktree_path, + "source_frozen": True, + "lock_record": lock_record, + "journal": journal, + "evidence": { + "written_paths": written, + "conflict_state_path": conflict_state_path, + "accepted_head": expected_remote_head, + "recovery_provenance": "dirty_orphan_recovery", + }, + } + + +def preflight_recognizes_recovered_provenance( + lock: Mapping[str, Any] | None, +) -> dict[str, Any]: + """Whether commit/publication preflights should accept recovered provenance.""" + if not lock: + return {"recognized": False, "reasons": ["no lock"]} + rec = lock.get("dirty_orphan_recovery") + if not isinstance(rec, Mapping) or not rec.get("recovered"): + return {"recognized": False, "reasons": ["no dirty_orphan_recovery record"]} + conflicts = rec.get("conflicts") or [] + if conflicts: + return { + "recognized": False, + "reasons": [ + "recovery conflicts remain; author must resolve before " + "commit/publication preflight" + ], + "conflicts": list(conflicts), + } + if not _text(lock.get("worktree_path")): + return {"recognized": False, "reasons": ["recovered lock missing worktree"]} + if _recorded_pid(lock) is None: + return { + "recognized": False, + "reasons": ["recovered lock still PID-less; binding incomplete"], + } + return { + "recognized": True, + "reasons": [], + "recovery_worktree_path": rec.get("recovery_worktree_path"), + "source_worktree_path": rec.get("source_worktree_path"), + "accepted_head": rec.get("accepted_head"), + } diff --git a/dirty_same_claimant_session_rebind.py b/dirty_same_claimant_session_rebind.py new file mode 100644 index 0000000..fc29add --- /dev/null +++ b/dirty_same_claimant_session_rebind.py @@ -0,0 +1,1565 @@ +"""Dirty-preserving same-claimant author-session rebind (#864 / #868). + +A registered issue worktree can be dirty while its durable lock owner PID is +provably dead. Ordinary ``gitea_lock_issue`` refuses dirty trees, and dead-session +recovery (#753) also requires cleanliness. This module is the *only* sanctioned +path that rebinds session/lock provenance onto the *same* worktree without +touching tracked or untracked content. + +This is SEPARATE from #860 dirty-orphan recovery (PID-less + remote sync). +This operation: + +* acts only on an already-registered dirty worktree +* updates only stale lock/session provenance (PID, session pointer, generation, + heartbeat) +* preserves every tracked/untracked byte +* does NOT sync remote, create recovery worktrees, clean, reset, or change heads + +#868 hardens: + +* complete dirty-inventory revalidation (full path set + fingerprints) + immediately before and after ``bind_session_lock`` +* durable recovery-journal operation identity (remote, org, repo, claimant + identity, claimant profile) validated on execute / resume / retry / + already_rebound +""" + +from __future__ import annotations + +import hashlib +import json +import os +import subprocess +import tempfile +from datetime import datetime, timezone +from typing import Any, Mapping, Sequence + +from author_mutation_worktree import is_path_under_branches +from issue_lock_provenance import ( + SOURCE_DIRTY_SAME_CLAIMANT_REBIND, + build_sanctioned_lock_provenance, +) +from issue_lock_store import ( + AUTHOR_ISSUE_WORK_LEASE, + bind_session_lock, + is_process_alive, + lock_file_path, + lock_generation, + read_lock_file, +) +from reviewer_worktree import parse_dirty_tracked_files + +# Outcomes +REBIND_SANCTIONED = "REBIND_SANCTIONED" +REFUSED = "REFUSED" +NO_CANDIDATE = "NO_CANDIDATE" + +# Provenance / tool identity +SOURCE_TOOL = SOURCE_DIRTY_SAME_CLAIMANT_REBIND +SOURCE = SOURCE_DIRTY_SAME_CLAIMANT_REBIND + +# Journal phases (crash-safe apply) +JOURNAL_PHASE_ASSESSED = "assessed" +JOURNAL_PHASE_PRE_BIND = "pre_bind" +JOURNAL_PHASE_BOUND = "bound" +JOURNAL_PHASE_COMPLETE = "complete" +JOURNAL_PHASE_ALREADY_REBOUND = "already_rebound" + +REQUIRED_LOCK_FIELDS = ( + "issue_number", + "branch_name", + "worktree_path", + "remote", + "org", + "repo", +) + +# Durable journal operation identity (#868 F2). All five must be persisted on +# JOURNAL_PHASE_ASSESSED and re-validated on resume / retry / already_rebound. +REQUIRED_JOURNAL_IDENTITY_FIELDS = ( + "remote", + "org", + "repo", + "claimant_identity", + "claimant_profile", +) + + +def _utc_now_iso() -> str: + return datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ") + + +def _text(value: Any) -> str: + return str(value or "").strip() + + +def _same_realpath(left: str | None, right: str | None) -> bool: + if not left or not right: + return False + try: + return os.path.realpath(left) == os.path.realpath(right) + except OSError: + return left == right + + +def _lock_claimant(lock: Mapping[str, Any]) -> dict[str, Any]: + claimant = lock.get("claimant") + if not isinstance(claimant, Mapping): + lease = lock.get("work_lease") + claimant = lease.get("claimant") if isinstance(lease, Mapping) else None + return dict(claimant) if isinstance(claimant, Mapping) else {} + + +def _recorded_pid(lock: Mapping[str, Any]) -> Any: + pid = lock.get("session_pid") + if pid is None: + pid = lock.get("pid") + return pid + + +def content_fingerprint(path: str) -> str: + """Return sha256 hex digest of file bytes at *path*. + + Missing or unreadable files raise ``OSError`` / ``FileNotFoundError`` so + callers fail closed rather than inventing an empty hash. + """ + digest = hashlib.sha256() + with open(path, "rb") as handle: + while True: + chunk = handle.read(1024 * 1024) + if not chunk: + break + digest.update(chunk) + return digest.hexdigest() + + +def parse_dirty_paths(porcelain: str) -> list[str]: + """Tracked + untracked paths from ``git status --porcelain -uall`` output.""" + paths: list[str] = [] + seen: set[str] = set() + for line in (porcelain or "").splitlines(): + if not line or len(line) < 4: + continue + if line.startswith("??"): + path = line[3:].strip() + else: + path = line[3:].strip() + if " -> " in path: + path = path.split(" -> ", 1)[1].strip() + if not path or path in seen: + continue + seen.add(path) + paths.append(path) + return paths + + +def collect_dirty_inventory(worktree_path: str) -> dict[str, Any]: + """Observe dirty tracked + untracked paths and content fingerprints. + + Uses ``git status --porcelain -uall`` so every untracked file is listed + individually (not collapsed into a directory). + """ + path = (worktree_path or "").strip() + if not path: + return { + "worktree_path": path, + "porcelain_status": "", + "dirty_paths": [], + "fingerprints": {}, + "ok": False, + "reasons": ["worktree path is empty"], + } + + status_res = subprocess.run( + ["git", "-C", path, "status", "--porcelain", "-uall"], + capture_output=True, + text=True, + check=False, + ) + if status_res.returncode != 0: + err = (status_res.stderr or status_res.stdout or "").strip() + return { + "worktree_path": path, + "porcelain_status": "", + "dirty_paths": [], + "fingerprints": {}, + "ok": False, + "reasons": [f"git status failed in '{path}': {err or 'unknown error'}"], + } + + porcelain = status_res.stdout or "" + dirty_paths = parse_dirty_paths(porcelain) + fingerprints: dict[str, str] = {} + reasons: list[str] = [] + for rel in dirty_paths: + abs_path = os.path.join(path, rel) + if os.path.isdir(abs_path) and not os.path.islink(abs_path): + # Directories appear only if git reported them; fingerprinting a + # directory is not defined — fail closed. + reasons.append(f"dirty path '{rel}' is a directory; cannot fingerprint") + continue + try: + fingerprints[rel] = content_fingerprint(abs_path) + except OSError as exc: + reasons.append(f"could not fingerprint '{rel}': {exc}") + + return { + "worktree_path": os.path.realpath(path), + "porcelain_status": porcelain, + "dirty_paths": dirty_paths, + "fingerprints": fingerprints, + "ok": not reasons, + "reasons": reasons, + "tracked_dirty": parse_dirty_tracked_files(porcelain), + } + + +def revalidate_complete_dirty_inventory( + worktree_path: str, + *, + expected_dirty_paths: Sequence[str] | None, + expected_fingerprints: Mapping[str, str] | None, + phase: str = "inventory", +) -> dict[str, Any]: + """Collect the full dirty inventory and require exact pin equality (#868 F1). + + Unlike fingerprint-only checks over the expected path list, this recollects + the authoritative tracked+untracked inventory and refuses added, removed, + or renamed paths as well as fingerprint movement. + """ + reasons: list[str] = [] + inv = collect_dirty_inventory(worktree_path) + if inv.get("ok") is False: + reasons.extend(list(inv.get("reasons") or []) or [f"{phase}: dirty inventory collection failed"]) + + observed_paths = sorted( + {_text(p) for p in (inv.get("dirty_paths") or []) if _text(p)} + ) + pin_paths = sorted( + {_text(p) for p in (expected_dirty_paths or []) if _text(p)} + ) + if not pin_paths: + reasons.append( + f"{phase}: expected_dirty_paths pin is empty; complete inventory " + "revalidation requires a non-empty pin (fail closed)" + ) + if set(observed_paths) != set(pin_paths): + extra = sorted(set(observed_paths) - set(pin_paths)) + missing = sorted(set(pin_paths) - set(observed_paths)) + if extra: + reasons.append( + f"{phase}: complete dirty inventory path-set disagreement: " + f"unexpected paths {extra}" + ) + if missing: + reasons.append( + f"{phase}: complete dirty inventory path-set disagreement: " + f"missing expected paths {missing}" + ) + + obs_fps = { + _text(k): _text(v) + for k, v in dict(inv.get("fingerprints") or {}).items() + if _text(k) + } + pin_fps = { + _text(k): _text(v) + for k, v in dict(expected_fingerprints or {}).items() + if _text(k) + } + if not pin_fps: + reasons.append( + f"{phase}: expected_fingerprints pin is empty; byte-level pins " + "are required (fail closed)" + ) + else: + for rel, expected_hash in pin_fps.items(): + if rel not in set(pin_paths): + reasons.append( + f"{phase}: expected_fingerprints contains '{rel}' which is " + "not in expected_dirty_paths" + ) + continue + actual_hash = obs_fps.get(rel) + if not actual_hash: + reasons.append( + f"{phase}: fingerprint missing for dirty path '{rel}'" + ) + elif actual_hash != expected_hash: + reasons.append( + f"{phase}: fingerprint disagreement for '{rel}': " + f"observed {actual_hash}, expected {expected_hash}" + ) + for rel in observed_paths: + if rel not in pin_fps: + reasons.append( + f"{phase}: observed dirty path '{rel}' has no fingerprint pin" + ) + + return { + "ok": not reasons, + "reasons": reasons, + "inventory": inv, + "observed_dirty_paths": observed_paths, + "expected_dirty_paths": pin_paths, + "observed_fingerprints": obs_fps, + "expected_fingerprints": pin_fps, + "phase": phase, + } + + +def build_journal_operation_identity( + *, + remote: str, + org: str, + repo: str, + claimant_identity: str | None, + claimant_profile: str | None, +) -> dict[str, str]: + """Return the five-field durable operation identity for the recovery journal.""" + return { + "remote": _text(remote), + "org": _text(org), + "repo": _text(repo), + "claimant_identity": _text(claimant_identity), + "claimant_profile": _text(claimant_profile), + } + + +def validate_journal_operation_identity( + journal: Mapping[str, Any] | None, + *, + remote: str, + org: str, + repo: str, + claimant_identity: str | None, + claimant_profile: str | None, + require_present: bool = True, +) -> list[str]: + """Validate durable journal identity fields (#868 F2). + + Rejects missing, mismatched, stale, cross-repository, or cross-claimant + journal state. When *require_present* is True, incomplete legacy journals + (any of the five fields absent/empty) fail closed. + """ + reasons: list[str] = [] + if not isinstance(journal, Mapping): + if require_present: + reasons.append( + "recovery journal is missing or unreadable; complete operation " + "identity cannot be proven (fail closed)" + ) + return reasons + + expected = build_journal_operation_identity( + remote=remote, + org=org, + repo=repo, + claimant_identity=claimant_identity, + claimant_profile=claimant_profile, + ) + for field in REQUIRED_JOURNAL_IDENTITY_FIELDS: + observed = _text(journal.get(field)) + want = expected[field] + if not observed: + reasons.append( + f"recovery journal omits operation identity field '{field}' " + "(incomplete legacy or malformed journal identity; fail closed)" + ) + continue + if not want: + reasons.append( + f"caller pin for journal identity field '{field}' is empty " + "(fail closed)" + ) + continue + if observed != want: + reasons.append( + f"recovery journal identity mismatch for '{field}': " + f"journal={observed!r}, expected={want!r} " + "(cross-repository / cross-claimant / replay refused)" + ) + return reasons + + +def journal_path(lock_dir: str, issue_number: int) -> str: + root = (lock_dir or "").strip() + return os.path.join(root, f".rebind-journal-{int(issue_number)}.json") + + +def _atomic_write_json(path: str, data: dict[str, Any]) -> None: + parent = os.path.dirname(path) or "." + os.makedirs(parent, mode=0o700, exist_ok=True) + payload = json.dumps(data, indent=2, sort_keys=True) + "\n" + fd, temp_path = tempfile.mkstemp(prefix=".rebind-j-", suffix=".json", dir=parent) + try: + with os.fdopen(fd, "w", encoding="utf-8") as handle: + handle.write(payload) + handle.flush() + os.fsync(handle.fileno()) + os.replace(temp_path, path) + finally: + if os.path.exists(temp_path): + try: + os.remove(temp_path) + except OSError: + pass + + +def _read_json(path: str) -> dict[str, Any] | None: + if not path or not os.path.exists(path): + return None + try: + with open(path, encoding="utf-8") as handle: + data = json.load(handle) + except (OSError, json.JSONDecodeError): + return None + return data if isinstance(data, dict) else None + + +def _malformed_lock_reasons(lock: Mapping[str, Any]) -> list[str]: + missing: list[str] = [] + for field in REQUIRED_LOCK_FIELDS: + if not _text(lock.get(field)): + missing.append(field) + pid = _recorded_pid(lock) + if pid is None or _text(pid) == "": + missing.append("session_pid/pid") + else: + try: + if int(pid) <= 0: + missing.append("session_pid/pid") + except (TypeError, ValueError): + missing.append("session_pid/pid") + return missing + + +def _canonical_under_branches(worktree_path: str, repo_root: str | None) -> tuple[bool, list[str]]: + """Prove worktree is a realpath under ``/branches/`` with no symlink escape.""" + reasons: list[str] = [] + path = (worktree_path or "").strip() + if not path: + return False, ["worktree path is empty"] + try: + real = os.path.realpath(path) + except OSError as exc: + return False, [f"worktree path could not be realpath-resolved: {exc}"] + if not os.path.isdir(real): + reasons.append(f"worktree path '{path}' is not an existing directory") + + root = (repo_root or "").strip() + if root: + try: + root_real = os.path.realpath(root) + except OSError as exc: + return False, [f"repo root could not be realpath-resolved: {exc}"] + if not is_path_under_branches(real, root_real): + reasons.append( + f"worktree '{real}' is not under branches/ of repo root '{root_real}' " + "(unregistered/noncanonical worktree; fail closed)" + ) + # Symlink escape: the declared path must not resolve outside branches/. + declared_abs = os.path.abspath(path) + if os.path.islink(path) or declared_abs != real: + if not is_path_under_branches(real, root_real): + reasons.append( + f"worktree path '{path}' escapes branches/ via symlink/realpath " + f"(resolves to '{real}')" + ) + else: + # Without an explicit repo root, still require a /branches/ segment. + if not is_path_under_branches(real, None): + reasons.append( + f"worktree '{real}' is not under a branches/ directory " + "(unregistered/noncanonical worktree; fail closed)" + ) + return not reasons, reasons + + +def assess_dirty_same_claimant_session_rebind( + *, + remote: str, + org: str, + repo: str, + issue_number: int, + branch_name: str, + worktree_path: str, + claimant_identity: str | None, + claimant_profile: str | None, + old_pid: int | None, + expected_local_head: str | None, + expected_remote_head: str | None, + expected_dirty_paths: Sequence[str] | None, + expected_fingerprints: Mapping[str, str] | None, + existing_lock: Mapping[str, Any] | None, + current_identity: str | None, + current_profile: str | None, + role_kind: str | None, + current_pid: int | None, + current_branch: str | None, + local_head: str | None, + remote_head: str | None, + porcelain_status: str | None = None, + dirty_inventory: Mapping[str, Any] | None = None, + competing_live_locks: Sequence[Mapping[str, Any]] | None = None, + competing_sessions: Sequence[Mapping[str, Any]] | None = None, + workflow_lease_active: bool = False, + authorize_reconciler_execute: bool = False, + permission_allowed: bool = False, + repo_root: str | None = None, +) -> dict[str, Any]: + """Pure assessment: may this dirty same-claimant lock be session-rebound? + + Every pin must agree. ``permission_allowed=True`` alone is never ownership + proof. Fail closed on live old PID, foreign identity/profile, pin mismatch, + unregistered/noncanonical worktree, head movement, dirty path/fingerprint + disagreement, competing ownership, malformed lock, empty PID, wrong role. + """ + reasons: list[str] = [] + evidence: dict[str, Any] = { + "issue_number": issue_number, + "branch_name": branch_name, + "worktree_path": worktree_path, + "remote": remote, + "org": org, + "repo": repo, + "old_pid": old_pid, + "current_pid": current_pid if current_pid is not None else os.getpid(), + "role_kind": _text(role_kind).lower() or None, + "permission_allowed": bool(permission_allowed), + } + + if not existing_lock: + return _assessment_result( + NO_CANDIDATE, + False, + ["no existing durable lock for this issue; not a rebind candidate"], + evidence, + ) + + lock = dict(existing_lock) + if lock.get("issue_number") != issue_number: + return _assessment_result( + NO_CANDIDATE, + False, + [ + f"existing lock targets issue #{lock.get('issue_number')}, " + f"not #{issue_number}; not a rebind candidate" + ], + evidence, + ) + + missing = _malformed_lock_reasons(lock) + if missing: + return _assessment_result( + REFUSED, + False, + [ + "durable lock record is incomplete and cannot prove ownership " + f"(missing/unusable: {', '.join(missing)})" + ], + evidence, + ) + + recorded_pid = _recorded_pid(lock) + evidence["recorded_pid"] = recorded_pid + evidence["lock_generation"] = lock_generation(lock) + + # ── Role gate ─────────────────────────────────────────────────────────── + role = _text(role_kind).lower() + if role in {"reviewer", "merger"}: + reasons.append( + f"role '{role}' cannot rebind dirty same-claimant author sessions " + "(fail closed)" + ) + elif role == "reconciler": + if not authorize_reconciler_execute: + reasons.append( + "reconciler role requires authorize_reconciler_execute=True " + "to execute dirty same-claimant rebind (fail closed)" + ) + elif role == "author": + pass + elif role: + reasons.append(f"role '{role}' is not authorized for dirty same-claimant rebind") + else: + reasons.append("role_kind is unknown; dirty same-claimant rebind refused") + + # permission_allowed is explicitly NOT ownership proof + evidence["note_permission_not_ownership"] = ( + "permission_allowed is not treated as ownership proof" + ) + + # ── Repository / issue / branch / worktree pins ───────────────────────── + for field, expected in (("remote", remote), ("org", org), ("repo", repo)): + actual = _text(lock.get(field)) + if actual != _text(expected): + reasons.append( + f"lock {field} '{actual}' does not match requested '{_text(expected)}'" + ) + + locked_branch = _text(lock.get("branch_name")) + if locked_branch != _text(branch_name): + reasons.append( + f"lock branch '{locked_branch}' does not match requested " + f"'{_text(branch_name)}'" + ) + + checked_out = _text(current_branch) + if not checked_out: + reasons.append( + "worktree is not on a named branch (detached HEAD); locked-branch " + "occupancy could not be proven" + ) + elif checked_out != locked_branch: + reasons.append( + f"worktree is on branch '{checked_out}', not the locked branch " + f"'{locked_branch}'" + ) + + locked_worktree = _text(lock.get("worktree_path")) + if not _same_realpath(locked_worktree, worktree_path): + reasons.append( + f"lock worktree '{locked_worktree}' does not match declared " + f"'{_text(worktree_path)}'" + ) + evidence["locked_worktree_path"] = locked_worktree + + under_ok, under_reasons = _canonical_under_branches(worktree_path, repo_root) + if not under_ok: + reasons.extend(under_reasons) + + # ── old_pid pin + liveness ────────────────────────────────────────────── + if old_pid is None or _text(old_pid) == "": + reasons.append("old_pid pin is empty; rebind refused (fail closed)") + else: + try: + old_pid_i = int(old_pid) + except (TypeError, ValueError): + reasons.append(f"old_pid '{old_pid}' is not a valid PID") + old_pid_i = None + if old_pid_i is not None: + if old_pid_i <= 0: + reasons.append("old_pid must be a positive integer (fail closed)") + try: + recorded_i = int(recorded_pid) + except (TypeError, ValueError): + recorded_i = None + if recorded_i is None or recorded_i != old_pid_i: + reasons.append( + f"old_pid {old_pid_i} does not match lock session_pid/pid " + f"{recorded_pid}" + ) + if is_process_alive(old_pid_i): + reasons.append( + f"old_pid {old_pid_i} is still alive; dirty same-claimant " + "rebind requires a provably dead owner (fail closed)" + ) + evidence["old_pid_alive"] = is_process_alive(old_pid_i) + if current_pid is not None: + try: + if int(current_pid) == old_pid_i: + reasons.append( + "old_pid is the current session PID; nothing to rebind" + ) + except (TypeError, ValueError): + pass + + # ── Claimant identity / profile ───────────────────────────────────────── + lock_claimant = _lock_claimant(lock) + locked_identity = _text(lock_claimant.get("username")) + locked_profile = _text(lock_claimant.get("profile")) + pin_identity = _text(claimant_identity) + pin_profile = _text(claimant_profile) + active_identity = _text(current_identity) + active_profile = _text(current_profile) + evidence["locked_identity"] = locked_identity or None + evidence["locked_profile"] = locked_profile or None + + if not locked_identity or not locked_profile: + reasons.append( + "durable lock does not record a claimant identity/profile; " + "ownership could not be proven" + ) + if not pin_identity or not pin_profile: + reasons.append( + "claimant_identity/claimant_profile pins are required (fail closed)" + ) + if locked_identity and pin_identity and locked_identity != pin_identity: + reasons.append( + f"claimant_identity pin '{pin_identity}' does not match lock " + f"claimant '{locked_identity}'" + ) + if locked_profile and pin_profile and locked_profile != pin_profile: + reasons.append( + f"claimant_profile pin '{pin_profile}' does not match lock profile " + f"'{locked_profile}'" + ) + + # Author path: active session must be the same claimant. Reconciler execute + # may rebind for the recorded claimant when explicitly authorized. + if role == "author": + if not active_identity or not active_profile: + reasons.append( + "active session identity/profile is unknown; author ownership " + "could not be proven" + ) + if locked_identity and active_identity and locked_identity != active_identity: + reasons.append( + f"lock claimant '{locked_identity}' does not match active " + f"identity '{active_identity}' (foreign claimant refused)" + ) + if locked_profile and active_profile and locked_profile != active_profile: + reasons.append( + f"lock profile '{locked_profile}' does not match active profile " + f"'{active_profile}' (profile mismatch refused)" + ) + if pin_identity and active_identity and pin_identity != active_identity: + reasons.append( + f"claimant_identity pin '{pin_identity}' does not match active " + f"identity '{active_identity}'" + ) + if pin_profile and active_profile and pin_profile != active_profile: + reasons.append( + f"claimant_profile pin '{pin_profile}' does not match active " + f"profile '{active_profile}'" + ) + + # ── Heads (must match pins and each other for this rebind class) ──────── + obs_local = _text(local_head) + obs_remote = _text(remote_head) + pin_local = _text(expected_local_head) + pin_remote = _text(expected_remote_head) + evidence["local_head"] = obs_local or None + evidence["remote_head"] = obs_remote or None + evidence["expected_local_head"] = pin_local or None + evidence["expected_remote_head"] = pin_remote or None + + if not pin_local or not pin_remote: + reasons.append( + "expected_local_head and expected_remote_head pins are required " + "(fail closed)" + ) + if not obs_local: + reasons.append("local head SHA could not be determined") + if not obs_remote: + reasons.append("remote head SHA could not be determined") + if pin_local and obs_local and pin_local != obs_local: + reasons.append( + f"local head moved or mismatched pin: observed {obs_local}, " + f"expected {pin_local}" + ) + if pin_remote and obs_remote and pin_remote != obs_remote: + reasons.append( + f"remote head moved or mismatched pin: observed {obs_remote}, " + f"expected {pin_remote}" + ) + if obs_local and obs_remote and obs_local != obs_remote: + # Dirty rebind does not allow unpublished head movement; heads must agree. + reasons.append( + f"local head {obs_local} does not match remote head {obs_remote}; " + "dirty same-claimant rebind requires matching heads (fail closed)" + ) + + # ── Dirty inventory + fingerprint pins ────────────────────────────────── + inv: dict[str, Any] + if isinstance(dirty_inventory, Mapping) and dirty_inventory.get("dirty_paths") is not None: + inv = dict(dirty_inventory) + if not inv.get("fingerprints") and porcelain_status is not None: + # Allow fingerprints-only refresh via recompute if needed. + pass + elif porcelain_status is not None: + # Porcelain alone proves path set, not bytes. Fingerprints must come from + # dirty_inventory (or apply()'s collect_dirty_inventory) — never from the + # caller's expected_fingerprints pin (that would make the pin tautological). + dirty_paths_obs = parse_dirty_paths(porcelain_status) + inv = { + "porcelain_status": porcelain_status, + "dirty_paths": dirty_paths_obs, + "fingerprints": {}, + "ok": True, + "reasons": [], + } + else: + reasons.append( + "neither dirty_inventory nor porcelain_status was provided; " + "dirty state could not be proven" + ) + inv = {"dirty_paths": [], "fingerprints": {}, "ok": False} + + if inv.get("ok") is False and inv.get("reasons"): + reasons.extend(list(inv.get("reasons") or [])) + + observed_paths = sorted({_text(p) for p in (inv.get("dirty_paths") or []) if _text(p)}) + pin_paths = sorted({_text(p) for p in (expected_dirty_paths or []) if _text(p)}) + evidence["observed_dirty_paths"] = observed_paths + evidence["expected_dirty_paths"] = pin_paths + + if not pin_paths: + reasons.append( + "expected_dirty_paths pin is empty; dirty same-claimant rebind " + "requires a non-empty dirty inventory pin (fail closed)" + ) + if set(observed_paths) != set(pin_paths): + extra = sorted(set(observed_paths) - set(pin_paths)) + missing_p = sorted(set(pin_paths) - set(observed_paths)) + if extra: + reasons.append( + f"dirty path set disagreement: unexpected paths {extra}" + ) + if missing_p: + reasons.append( + f"dirty path set disagreement: missing expected paths {missing_p}" + ) + + obs_fps = { + _text(k): _text(v) + for k, v in dict(inv.get("fingerprints") or {}).items() + if _text(k) + } + pin_fps = { + _text(k): _text(v) + for k, v in dict(expected_fingerprints or {}).items() + if _text(k) + } + evidence["observed_fingerprints"] = obs_fps + evidence["expected_fingerprints"] = pin_fps + + if not pin_fps: + reasons.append( + "expected_fingerprints pin is empty; byte-level pins are required " + "(fail closed)" + ) + else: + for rel, expected_hash in pin_fps.items(): + if rel not in set(pin_paths): + reasons.append( + f"expected_fingerprints contains '{rel}' which is not in " + "expected_dirty_paths" + ) + actual_hash = obs_fps.get(rel) + if not actual_hash: + reasons.append( + f"fingerprint missing for dirty path '{rel}'" + ) + elif actual_hash != expected_hash: + reasons.append( + f"fingerprint disagreement for '{rel}': observed " + f"{actual_hash}, expected {expected_hash}" + ) + for rel in obs_fps: + if rel in set(pin_paths) and rel not in pin_fps: + reasons.append( + f"expected_fingerprints missing pin for observed dirty path '{rel}'" + ) + + # ── Competing ownership ───────────────────────────────────────────────── + competing: list[dict[str, Any]] = [] + for entry in competing_live_locks or (): + if not isinstance(entry, Mapping): + continue + same_issue = entry.get("issue_number") == issue_number + same_branch = _text(entry.get("branch_name")) == locked_branch + if not (same_issue or same_branch): + continue + if ( + same_issue + and same_branch + and _same_realpath(_text(entry.get("worktree_path")), worktree_path) + ): + # The lock we are rebinding is not competition with itself, but a + # *live* competing owner on the same worktree is still a problem. + entry_pid = entry.get("pid") or entry.get("session_pid") + try: + entry_pid_i = int(entry_pid) if entry_pid is not None else None + except (TypeError, ValueError): + entry_pid_i = None + if entry_pid_i is not None and is_process_alive(entry_pid_i): + if old_pid is None or entry_pid_i != int(old_pid): + competing.append( + { + "issue_number": entry.get("issue_number"), + "branch_name": entry.get("branch_name"), + "worktree_path": entry.get("worktree_path"), + "pid": entry_pid_i, + } + ) + continue + competing.append( + { + "issue_number": entry.get("issue_number"), + "branch_name": entry.get("branch_name"), + "worktree_path": entry.get("worktree_path"), + "pid": entry.get("pid") or entry.get("session_pid"), + } + ) + if competing: + described = ", ".join( + f"issue #{c['issue_number']} branch '{c['branch_name']}' pid={c.get('pid')}" + for c in competing + ) + reasons.append(f"competing live lock exists ({described})") + evidence["competing_live_locks"] = competing + + competing_sess: list[dict[str, Any]] = [] + for entry in competing_sessions or (): + if not isinstance(entry, Mapping): + continue + sess_pid = entry.get("pid") or entry.get("session_pid") + try: + sess_pid_i = int(sess_pid) if sess_pid is not None else None + except (TypeError, ValueError): + sess_pid_i = None + if sess_pid_i is None: + continue + if current_pid is not None and sess_pid_i == int(current_pid): + continue + if old_pid is not None: + try: + if sess_pid_i == int(old_pid) and not is_process_alive(sess_pid_i): + continue + except (TypeError, ValueError): + pass + if is_process_alive(sess_pid_i) or entry.get("live") is True: + competing_sess.append( + { + "pid": sess_pid_i, + "lock_file_path": entry.get("lock_file_path"), + } + ) + if competing_sess: + reasons.append( + "competing live session pointer(s) claim this lock: " + + ", ".join(str(s["pid"]) for s in competing_sess) + ) + evidence["competing_sessions"] = competing_sess + + if workflow_lease_active: + reasons.append( + "workflow lease is active for this scope; dirty same-claimant " + "rebind refused (fail closed)" + ) + evidence["workflow_lease_active"] = bool(workflow_lease_active) + + if reasons: + return _assessment_result(REFUSED, False, reasons, evidence) + + proof = [ + f"registered dirty worktree for issue #{issue_number} on branch " + f"'{locked_branch}' matches claimant '{locked_identity}' / profile " + f"'{locked_profile}'; old_pid {recorded_pid} is dead; heads " + f"{obs_local} match; {len(pin_paths)} dirty paths fingerprint-pinned; " + "provenance-only rebind sanctioned" + ] + return _assessment_result(REBIND_SANCTIONED, True, proof, evidence) + + +def _assessment_result( + outcome: str, + sanctioned: bool, + reasons: list[str], + evidence: dict[str, Any], +) -> dict[str, Any]: + return { + "outcome": outcome, + "rebind_sanctioned": sanctioned, + "is_candidate": outcome != NO_CANDIDATE, + "reasons": reasons, + "evidence": evidence, + "expected_generation": evidence.get("lock_generation"), + } + + +def _already_rebound( + *, + existing_lock: Mapping[str, Any], + current_pid: int, + worktree_path: str, + expected_dirty_paths: Sequence[str] | None, + expected_fingerprints: Mapping[str, str], + worktree_for_fps: str, + remote: str, + org: str, + repo: str, + claimant_identity: str | None, + claimant_profile: str | None, + journal: Mapping[str, Any] | None = None, +) -> tuple[bool, list[str]]: + """Return (True, notes) when lock is already rebound to this session. + + #868: require complete matching operation identity (remote/org/repo/ + claimant) and complete dirty-inventory revalidation, not fingerprint-only + checks. Incomplete or mismatched journal identity fails closed. + """ + notes: list[str] = [] + pid = _recorded_pid(existing_lock) + try: + pid_i = int(pid) if pid is not None else None + except (TypeError, ValueError): + return False, [] + if pid_i != int(current_pid): + return False, [] + if not _same_realpath(_text(existing_lock.get("worktree_path")), worktree_path): + return False, [] + + # Durable lock repo binding must still match the caller's target. + for field, expected in (("remote", remote), ("org", org), ("repo", repo)): + observed = _text(existing_lock.get(field)) + want = _text(expected) + if observed and want and observed != want: + notes.append( + f"already_rebound refused: lock {field}={observed!r} does not " + f"match expected {want!r} (cross-repository replay)" + ) + return False, notes + + lock_claimant = _lock_claimant(existing_lock) + locked_identity = _text(lock_claimant.get("username")) + locked_profile = _text(lock_claimant.get("profile")) + pin_identity = _text(claimant_identity) + pin_profile = _text(claimant_profile) + if pin_identity and locked_identity and pin_identity != locked_identity: + notes.append( + f"already_rebound refused: lock claimant '{locked_identity}' does " + f"not match pin '{pin_identity}' (cross-claimant replay)" + ) + return False, notes + if pin_profile and locked_profile and pin_profile != locked_profile: + notes.append( + f"already_rebound refused: lock profile '{locked_profile}' does " + f"not match pin '{pin_profile}' (cross-claimant replay)" + ) + return False, notes + + # When a durable journal is present, require complete matching identity. + if isinstance(journal, Mapping) and journal: + id_reasons = validate_journal_operation_identity( + journal, + remote=remote, + org=org, + repo=repo, + claimant_identity=claimant_identity, + claimant_profile=claimant_profile, + require_present=True, + ) + if id_reasons: + notes.extend(id_reasons) + return False, notes + + inv_check = revalidate_complete_dirty_inventory( + worktree_for_fps, + expected_dirty_paths=expected_dirty_paths, + expected_fingerprints=expected_fingerprints, + phase="already_rebound", + ) + if not inv_check["ok"]: + notes.extend(list(inv_check["reasons"] or [])) + return False, notes + + gen = lock_generation(existing_lock) + if gen < 1: + # A never-written generation is suspicious for a completed rebind, but + # a same-pid lock with matching fingerprints is still "ours". + notes.append("lock generation is 0; treating same-pid match as rebound") + return True, notes or ["lock already bound to current session PID"] + + +def apply_dirty_same_claimant_session_rebind( + *, + remote: str, + org: str, + repo: str, + issue_number: int, + branch_name: str, + worktree_path: str, + claimant_identity: str | None, + claimant_profile: str | None, + old_pid: int | None, + expected_local_head: str | None, + expected_remote_head: str | None, + expected_dirty_paths: Sequence[str] | None, + expected_fingerprints: Mapping[str, str] | None, + existing_lock: Mapping[str, Any] | None, + current_identity: str | None, + current_profile: str | None, + role_kind: str | None, + current_pid: int | None = None, + current_branch: str | None, + local_head: str | None, + remote_head: str | None, + porcelain_status: str | None = None, + dirty_inventory: Mapping[str, Any] | None = None, + competing_live_locks: Sequence[Mapping[str, Any]] | None = None, + competing_sessions: Sequence[Mapping[str, Any]] | None = None, + workflow_lease_active: bool = False, + authorize_reconciler_execute: bool = False, + permission_allowed: bool = False, + repo_root: str | None = None, + dry_run: bool = False, + lock_dir: str | None = None, +) -> dict[str, Any]: + """Assess and (unless dry_run) apply a dirty same-claimant session rebind.""" + pid_now = int(current_pid) if current_pid is not None else os.getpid() + wt = os.path.realpath((worktree_path or "").strip()) if worktree_path else "" + + # Prefer a live inventory when applying so fingerprints are re-observed. + inv = dict(dirty_inventory) if isinstance(dirty_inventory, Mapping) else None + if inv is None and wt: + inv = collect_dirty_inventory(wt) + + assessment = assess_dirty_same_claimant_session_rebind( + remote=remote, + org=org, + repo=repo, + issue_number=issue_number, + branch_name=branch_name, + worktree_path=worktree_path, + claimant_identity=claimant_identity, + claimant_profile=claimant_profile, + old_pid=old_pid, + expected_local_head=expected_local_head, + expected_remote_head=expected_remote_head, + expected_dirty_paths=expected_dirty_paths, + expected_fingerprints=expected_fingerprints, + existing_lock=existing_lock, + current_identity=current_identity, + current_profile=current_profile, + role_kind=role_kind, + current_pid=pid_now, + current_branch=current_branch, + local_head=local_head, + remote_head=remote_head, + porcelain_status=porcelain_status + if porcelain_status is not None + else (inv or {}).get("porcelain_status"), + dirty_inventory=inv, + competing_live_locks=competing_live_locks, + competing_sessions=competing_sessions, + workflow_lease_active=workflow_lease_active, + authorize_reconciler_execute=authorize_reconciler_execute, + permission_allowed=permission_allowed, + repo_root=repo_root, + ) + + base_result: dict[str, Any] = { + "success": False, + "dry_run": bool(dry_run), + "outcome": assessment["outcome"], + "rebind_sanctioned": assessment["rebind_sanctioned"], + "reasons": list(assessment.get("reasons") or []), + "evidence": assessment.get("evidence") or {}, + "old_pid": old_pid, + "new_pid": pid_now, + "already_rebound": False, + "dirty_paths": list(expected_dirty_paths or []), + "fingerprints": dict(expected_fingerprints or {}), + "local_head": local_head, + "remote_head": remote_head, + } + + root_for_journal = (lock_dir or "").strip() or None + if root_for_journal is None and isinstance(existing_lock, Mapping): + root_for_journal = os.path.dirname( + _text(existing_lock.get("lock_file_path")) + or lock_file_path( + remote=remote, org=org, repo=repo, issue_number=issue_number + ) + ) + jpath_probe = ( + journal_path(root_for_journal, issue_number) if root_for_journal else "" + ) + existing_journal = _read_json(jpath_probe) if jpath_probe else None + + # Resume / retry: reject incomplete, mismatched, or cross-repo journal + # identity before treating any prior journal as authoritative (#868 F2). + if isinstance(existing_journal, Mapping) and existing_journal: + journal_id_reasons = validate_journal_operation_identity( + existing_journal, + remote=remote, + org=org, + repo=repo, + claimant_identity=claimant_identity, + claimant_profile=claimant_profile, + require_present=True, + ) + # Incomplete legacy journals from pre-#868 apply paths must fail closed + # when any identity field is missing — even if the rest of the payload + # looks familiar. Only a complete matching identity may proceed. + phase = _text(existing_journal.get("phase")) + if journal_id_reasons and phase not in ("", JOURNAL_PHASE_COMPLETE): + # Allow a completed journal with missing legacy identity only when + # already_rebound path will re-validate lock + inventory; for + # mid-flight incomplete journals, refuse. + if phase in ( + JOURNAL_PHASE_ASSESSED, + JOURNAL_PHASE_PRE_BIND, + JOURNAL_PHASE_BOUND, + "bind_failed", + ): + return { + **base_result, + "success": False, + "reasons": journal_id_reasons, + "journal_path": jpath_probe, + "journal_phase": phase or None, + } + + # Retry-safe: if already rebound to this session, succeed even when assess + # refuses because old_pid no longer matches the (updated) lock. + if ( + isinstance(existing_lock, Mapping) + and expected_fingerprints + and wt + ): + done, notes = _already_rebound( + existing_lock=existing_lock, + current_pid=pid_now, + worktree_path=worktree_path, + expected_dirty_paths=expected_dirty_paths, + expected_fingerprints=expected_fingerprints, + worktree_for_fps=wt, + remote=remote, + org=org, + repo=repo, + claimant_identity=claimant_identity, + claimant_profile=claimant_profile, + journal=existing_journal, + ) + if done: + lock_path = _text(existing_lock.get("lock_file_path")) or lock_file_path( + remote=remote, + org=org, + repo=repo, + issue_number=issue_number, + lock_dir=lock_dir, + ) + session_ptr = os.path.join( + (lock_dir or os.path.dirname(lock_path) or "."), + f"session-{pid_now}.json", + ) + return { + **base_result, + "success": True, + "outcome": REBIND_SANCTIONED, + "rebind_sanctioned": True, + "already_rebound": True, + "reasons": notes, + "lock_path": lock_path, + "session_pointer": session_ptr, + "generation_before": lock_generation(existing_lock), + "generation_after": lock_generation(existing_lock), + "journal_phase": JOURNAL_PHASE_ALREADY_REBOUND, + } + # Same-pid candidate that failed complete identity/inventory checks + # must not fall through into a fresh bind that would re-mint authority. + if _recorded_pid(existing_lock) is not None: + try: + if int(_recorded_pid(existing_lock)) == int(pid_now) and notes: + return { + **base_result, + "success": False, + "already_rebound": False, + "reasons": notes, + "journal_path": jpath_probe or None, + } + except (TypeError, ValueError): + pass + + if not assessment["rebind_sanctioned"]: + return base_result + + if dry_run: + return { + **base_result, + "success": True, + "message": "dry_run: rebind sanctioned; no lock/session writes performed", + "generation_before": assessment.get("expected_generation"), + "generation_after": assessment.get("expected_generation"), + } + + lock = dict(existing_lock or {}) + gen_before = lock_generation(lock) + root = (lock_dir or "").strip() or None + jpath = journal_path( + root or os.path.dirname( + _text(lock.get("lock_file_path")) + or lock_file_path( + remote=remote, org=org, repo=repo, issue_number=issue_number + ) + ), + issue_number, + ) + + op_identity = build_journal_operation_identity( + remote=remote, + org=org, + repo=repo, + claimant_identity=claimant_identity, + claimant_profile=claimant_profile, + ) + # Refuse incomplete caller identity before any durable write. + for field, value in op_identity.items(): + if not value: + return { + **base_result, + "success": False, + "reasons": [ + f"cannot write recovery journal: operation identity field " + f"'{field}' is empty (fail closed)" + ], + } + + journal = { + "phase": JOURNAL_PHASE_ASSESSED, + "issue_number": issue_number, + "branch_name": branch_name, + "worktree_path": wt, + "old_pid": old_pid, + "new_pid": pid_now, + "expected_generation": gen_before, + "expected_fingerprints": dict(expected_fingerprints or {}), + "expected_dirty_paths": list(expected_dirty_paths or []), + "local_head": local_head, + "remote_head": remote_head, + "started_at": _utc_now_iso(), + "source": SOURCE, + # #868 F2 — complete durable operation identity + **op_identity, + } + _atomic_write_json(jpath, journal) + + # #868 F1 — complete dirty-inventory revalidation immediately before mutation. + # Fail closed with no bind so failures cannot leave a newly authoritative + # live session. + pre_inv = revalidate_complete_dirty_inventory( + wt, + expected_dirty_paths=expected_dirty_paths, + expected_fingerprints=expected_fingerprints, + phase="pre-bind", + ) + if not pre_inv["ok"]: + journal["phase"] = "pre_bind_inventory_failed" + journal["pre_bind_inventory"] = { + "observed_dirty_paths": pre_inv.get("observed_dirty_paths"), + "reasons": pre_inv.get("reasons"), + } + _atomic_write_json(jpath, journal) + return { + **base_result, + "success": False, + "reasons": list(pre_inv["reasons"] or []), + "journal_path": jpath, + "journal_phase": "pre_bind_inventory_failed", + } + + pre_fps = dict(pre_inv.get("observed_fingerprints") or {}) + + journal["phase"] = JOURNAL_PHASE_PRE_BIND + journal["pre_bind_fingerprints"] = pre_fps + journal["pre_bind_dirty_paths"] = list(pre_inv.get("observed_dirty_paths") or []) + _atomic_write_json(jpath, journal) + + now = _utc_now_iso() + new_lock = dict(lock) + new_lock["session_pid"] = pid_now + new_lock["pid"] = pid_now + new_lock["last_heartbeat_at"] = now + new_lock["remote"] = remote + new_lock["org"] = org + new_lock["repo"] = repo + new_lock["issue_number"] = issue_number + new_lock["branch_name"] = branch_name + new_lock["worktree_path"] = _text(lock.get("worktree_path")) or wt + + # Preserve work_lease (including expires_at); refresh heartbeat only. + lease = new_lock.get("work_lease") + if isinstance(lease, dict): + lease = dict(lease) + lease["last_heartbeat_at"] = now + if not lease.get("operation_type"): + lease["operation_type"] = AUTHOR_ISSUE_WORK_LEASE + new_lock["work_lease"] = lease + + claimant = _lock_claimant(lock) + new_lock["lock_provenance"] = build_sanctioned_lock_provenance( + tool=SOURCE_TOOL, + source=SOURCE, + claimant=claimant or { + "username": claimant_identity, + "profile": claimant_profile, + }, + ) + new_lock["rebind_record"] = { + "source": SOURCE, + "old_pid": old_pid, + "new_pid": pid_now, + "rebound_at": now, + "local_head": local_head, + "remote_head": remote_head, + "dirty_path_count": len(list(expected_dirty_paths or [])), + "generation_before": gen_before, + "evidence": { + "fingerprints": dict(expected_fingerprints or {}), + "dirty_paths": list(expected_dirty_paths or []), + }, + } + + try: + lock_path = bind_session_lock( + new_lock, + lock_dir=root, + expected_generation=gen_before, + ) + except Exception as exc: + journal["phase"] = "bind_failed" + journal["error"] = str(exc) + _atomic_write_json(jpath, journal) + return { + **base_result, + "success": False, + "reasons": [f"bind_session_lock failed: {exc}"], + "journal_path": jpath, + "generation_before": gen_before, + } + + journal["phase"] = JOURNAL_PHASE_BOUND + journal["lock_path"] = lock_path + _atomic_write_json(jpath, journal) + + # #868 F1 — complete dirty-inventory revalidation immediately after mutation. + # Path set must remain exactly equal; fingerprints must be unchanged. + post_inv = revalidate_complete_dirty_inventory( + wt, + expected_dirty_paths=expected_dirty_paths, + expected_fingerprints=expected_fingerprints, + phase="post-bind", + ) + post_fps = dict(post_inv.get("observed_fingerprints") or {}) + if not post_inv["ok"]: + journal["phase"] = "post_bind_inventory_failed" + journal["post_bind_inventory"] = { + "observed_dirty_paths": post_inv.get("observed_dirty_paths"), + "reasons": post_inv.get("reasons"), + } + journal["post_bind_fingerprints"] = post_fps + _atomic_write_json(jpath, journal) + return { + **base_result, + "success": False, + "reasons": list(post_inv["reasons"] or []) + [ + "post-bind complete inventory revalidation failed after " + "bind_session_lock; lock may be rebound but content/path " + "verification failed (fail closed)" + ], + "lock_path": lock_path, + "journal_path": jpath, + "journal_phase": "post_bind_inventory_failed", + "generation_before": gen_before, + "fingerprints_after": post_fps, + } + + # Remove stale session pointer for old_pid when it points at this lock. + removed_old_pointer = False + if old_pid is not None and root: + old_ptr = os.path.join(root, f"session-{int(old_pid)}.json") + if os.path.exists(old_ptr): + ptr = _read_json(old_ptr) or {} + ptr_lock = _text(ptr.get("lock_file_path")) + if not ptr_lock or os.path.realpath(ptr_lock) == os.path.realpath(lock_path): + try: + os.remove(old_ptr) + removed_old_pointer = True + except OSError: + pass + + bound = read_lock_file(lock_path) or new_lock + gen_after = lock_generation(bound) + session_ptr = os.path.join( + root or os.path.dirname(lock_path), + f"session-{pid_now}.json", + ) + + journal["phase"] = JOURNAL_PHASE_COMPLETE + journal["completed_at"] = _utc_now_iso() + journal["generation_after"] = gen_after + journal["removed_old_session_pointer"] = removed_old_pointer + journal["post_bind_fingerprints"] = post_fps + journal["post_bind_dirty_paths"] = list( + post_inv.get("observed_dirty_paths") or [] + ) + _atomic_write_json(jpath, journal) + + return { + **base_result, + "success": True, + "message": ( + f"Rebound dirty same-claimant author session for issue #{issue_number} " + f"from dead pid {old_pid} to pid {pid_now}; dirty bytes preserved" + ), + "lock_path": lock_path, + "session_pointer": session_ptr, + "generation_before": gen_before, + "generation_after": gen_after, + "fingerprints": post_fps, + "fingerprints_before": pre_fps, + "removed_old_session_pointer": removed_old_pointer, + "journal_path": jpath, + "journal_phase": JOURNAL_PHASE_COMPLETE, + "rebind_record": bound.get("rebind_record") or new_lock.get("rebind_record"), + "lock_provenance": bound.get("lock_provenance"), + } + + +def build_issue_860_regression_fixture_spec() -> dict[str, Any]: + """Data-only fixture describing the #860 class scenario (no real mutation). + + Claimant jcwalker3 / prgs-author, dead PID, no live session pointer, seven + dirty paths with fingerprint pins, matching local/remote heads. + """ + dirty_paths = [ + "dirty_same_claimant_session_rebind.py", + "issue_lock_provenance.py", + "task_capability_map.py", + "gitea_mcp_server.py", + "tests/test_dirty_same_claimant_session_rebind.py", + "docs/runbook-dirty-rebind.md", + "scratch/notes-untracked.txt", + ] + # Stable placeholder digests — tests replace with real fingerprints when + # constructing on-disk fixtures. These exist so the spec is self-describing. + fingerprints = { + path: hashlib.sha256(f"issue-860-fixture:{path}".encode()).hexdigest() + for path in dirty_paths + } + head = "a" * 40 + dead = 424860 + return { + "issue_class": "issue-860-dirty-orphan-class-fixture", + "description": ( + "Registered dirty worktree, same claimant, dead owner PID, no live " + "session pointer, seven fingerprint-pinned dirty paths, matching heads. " + "Data only — does not mutate any real worktree." + ), + "remote": "prgs", + "org": "Scaled-Tech-Consulting", + "repo": "Gitea-Tools", + "issue_number": 860, + "branch_name": "fix/issue-860-dirty-orphan-recovery", + "worktree_path": "/scratch/branches/fix-issue-860-dirty-orphan-recovery", + "claimant_identity": "jcwalker3", + "claimant_profile": "prgs-author", + "old_pid": dead, + "old_pid_alive": False, + "live_session_pointer": None, + "expected_local_head": head, + "expected_remote_head": head, + "expected_dirty_paths": dirty_paths, + "expected_fingerprints": fingerprints, + "dirty_path_count": 7, + "role_kind": "author", + "notes": [ + "Distinct from #864 apply path: this fixture documents the #860 class " + "inputs (dead PID + dirty inventory) without remote sync or recovery " + "worktree creation.", + ], + } diff --git a/docs/architecture/mcp-restart-governance.md b/docs/architecture/mcp-restart-governance.md new file mode 100644 index 0000000..2931d10 --- /dev/null +++ b/docs/architecture/mcp-restart-governance.md @@ -0,0 +1,223 @@ +# ADR: MCP restart governance and authorization policy + +- **Status:** Accepted (policy effective immediately for LLM and operator sessions; enforcement tooling may lag) +- **Date:** 2026-07-23 +- **Tracking issue:** [#656](https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/issues/656) +- **Policy version:** `restart-governance/v1` +- **Related:** + - Umbrella: [#655](https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/issues/655) — governed MCP restart coordination and zero-disruption recovery + - Vision: [#652](https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/issues/652) — MCP Control Plane Web Console product vision (§A system health and process control) + - Roadmap: [#653](https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/issues/653) — Control Plane Web Console phased delivery (Phase 2 restart controls) + - Contamination guard: [#630](https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/issues/630) — blocks manual process-kill recovery + - Console restart UX: [#642](https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/issues/642) — sanctioned restart and graceful reload + - Existing restart / reconnect paths to inventory: [#591](https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/issues/591) — auto-restart on master advance (closed); [#584](https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/issues/584) — host auto-reconnect on transport flap + - Stable-control runtime split: `docs/architecture/mcp-stable-control-runtime-policy-adr.md` (#615) + - Client-namespace health: `docs/mcp-namespace-health.md` (#543) + - Reconnect-only EOF recovery: `docs/mcp-namespace-eof-recovery.md` + +## 1. Context + +The Gitea MCP server is the **control plane** for real issue and PR mutations +(create, comment, lock, review, merge, reconcile). The same process serves every +role namespace (`gitea-author`, `gitea-reviewer`, `gitea-merger`, +`gitea-reconciler`, `gitea-controller`) and holds the in-memory capability-gate +code loaded at startup. + +Restarting that process is destructive to concurrent work: + +- It resets every session's identity, preflight, and capability-lease binding. +- It can interrupt a mutation mid-critical-section (a lock acquire, a review + submit, a merge), leaving durable state half-written. +- Relaunching from the wrong checkout or worktree silently changes which code + the control plane runs, defeating master-parity gates (#420 / #615). + +Today there is **no durable written policy** stating who may restart MCP, under +what conditions, that restart is a last resort, and how controller approval, +automated safety gates, and break-glass interact. Operators and LLM sessions +therefore invent restart behavior ad hoc, which makes concurrent multi-role work +unsafe. #630 and #642 need this policy as their backbone. + +This ADR defines that policy. It does **not** implement coordinator code or HA +multi-instance restart (those are later children of #655). + +## 2. Decision + +### 2.1 v1 decision (recorded) + +**Restart authority in v1 is `controller approval + automated safety gates`.** + +A restart of the stable control runtime is authorized only when **both** hold: + +1. A **controller** role explicitly approves the restart, recording an audit + entry (who, why, scope, affected sessions), **and** +2. The **automated safety gates** pass: a completed drain acknowledgement (no + affected session is mid-critical-section) or a declared break-glass incident + (§2.5). + +Quorum among multiple controllers is **not** required day-one. It is deferred +unless a later investigation (tracked under #653) proves single-controller +approval is insufficient. This ADR records the v1 decision so enforcement code +(#630) has a fixed target; changing it requires a superseding ADR. + +### 2.2 Restart is a last resort — the recovery ladder + +Restart is the **last** rung. Before any restart, exhaust the narrower +recoveries, in order: + +1. **Reconnect** the IDE/client MCP namespace (transport EOF, `client is + closing: EOF`, transient `#584` flap). No process change. See + `docs/mcp-namespace-eof-recovery.md`. +2. **Refresh / rebind** the session workspace: re-run `gitea_whoami`, + `gitea_resolve_task_capability`, and pass an explicit validated + `worktree_path`. Fixes stale session context without touching the process. +3. **Scoped restart** of a single misbehaving namespace/service (where the + deployment supports per-service restart) rather than the whole control plane. +4. **Full restart** of the stable control runtime process — operator-owned, + controller-approved, drained. +5. **Host / infrastructure restart** — the broadest action; same authorization + as a full restart plus infrastructure ownership. + +A session **must** try rungs 1–2 and record why they were insufficient before +requesting a restart at rung 3 or above. Skipping straight to restart is a +policy violation. + +### 2.3 Authorization matrix + +| Role | Reconnect (1) | Refresh/rebind (2) | Scoped restart (3) | Full restart (4) | Host restart (5) | +|---|---|---|---|---|---| +| **author** | self | self | request only | **forbidden** | forbidden | +| **reviewer** | self | self | request only | **forbidden** | forbidden | +| **merger** | self | self | request only | **forbidden** | forbidden | +| **reconciler** | self | self | request only | **forbidden** | forbidden | +| **controller** | self | self | **approve** (+gates) | **approve** (+gates) | request to operator | +| **operator** | self | self | execute (controller-approved) | execute (controller-approved) | execute (controller-approved) | +| **admin** | self | self | execute | execute | execute (break-glass) | + +Legend: *self* = may perform for its own client session; *request only* = may +raise a restart request but not authorize or execute it; *approve* = may +authorize under §2.1 gates; *execute* = may perform the process action after the +authorization is recorded. + +Key invariants: + +- **No LLM worker role (author/reviewer/merger/reconciler) may perform or + authorize a full or host restart.** They may only reconnect/rebind their own + client and file a restart request. +- **Controller approval authorizes; operator/admin executes.** The approving + controller and the executing operator may be the same human, but both the + approval and the execution are audited. +- Privileged process actions (full restart, host restart) are reserved to + **operator/admin**, never to an automated worker. + +### 2.4 Approved conditions + +A restart at rung 3+ is approved only under one of these recorded conditions: + +- **No affected sessions:** the control plane has no live session that would be + interrupted (verified, not assumed). +- **Full drain acknowledged:** every affected session has drained + (no open critical section — no held mutation lease mid-write) and the drain is + acknowledged in the audit record. +- **Controller + gates:** controller approval plus passing automated safety + gates (§2.1), the standard v1 path. +- **Quorum:** not required in v1; reserved for a future superseding ADR. +- **Break-glass:** an incident-backed emergency exception (§2.5). + +Restart **never** bypasses mutation gates mid-critical-section. Drain before +restart is mandatory except under break-glass with a declared incident. + +### 2.5 Break-glass + +Break-glass is a **separate, narrower** authorization path for emergencies where +the normal drain-and-approve path cannot complete (e.g. the control plane is +wedged and cannot drain). + +Break-glass conditions: + +- A declared incident record exists (id, timestamp, declarer) **before** the + action. +- The action is taken by **operator or admin** authority only — never by an LLM + worker role, and never unilaterally by an operator with active peers when a + controller is reachable. +- The scope is the minimum necessary rung of the ladder. +- A **mandatory post-hoc audit** entry is filed: what was restarted, why the + normal path was impossible, which sessions were affected, and the incident id. + +Break-glass suspends the drain requirement, not the audit requirement. + +### 2.6 Explicit prohibitions + +- **A unilateral LLM or operator full restart while active peer sessions + exist is forbidden.** An LLM worker role must not kill, restart, or relaunch + the MCP process; a lone operator must not full-restart over live peer work + without controller approval or a break-glass incident. +- Process-kill recovery is forbidden as a routine tool (#630). This ADR does not + introduce a kill path. +- Ambiguous policy state **denies** restart (§4). + +## 3. Security requirements + +- Full restart and host restart are **privileged**; only operator/admin execute + them, only after a controller approval or break-glass incident is recorded. +- Break-glass is a distinct authorization path with its own audit mandate; it is + never the default and never silent. +- **Every approval and every restart action is audited** (who approved, who + executed, scope, affected sessions, condition, policy version). No restart is + authorized without a durable audit entry. + +## 4. Failure behavior + +**Ambiguous policy → deny restart.** If it cannot be established that a +restart is authorized under §2 — unknown affected-session state, missing +controller approval, absent break-glass incident, or an unclassifiable request — +the safe action is to **refuse** the restart and stop with a recovery report, +never to restart on assumption. + +## 5. Policy IDs (for enforcement code) + +Enforcement code — the restart coordinator (a later child of #655), the #630 +contamination guard, and the #642 console restart UX — binds to these stable +policy identifiers rather than to prose: + +| Policy ID | Statement | +|---|---| +| `RG-01` | Restart is last resort; rungs 1–2 must be tried and recorded first (§2.2). | +| `RG-02` | v1 authority = controller approval + automated safety gates (§2.1). | +| `RG-03` | No LLM worker role performs or authorizes full/host restart (§2.3). | +| `RG-04` | Full/host restart executed by operator/admin only, post approval (§2.3). | +| `RG-05` | Drain before restart is mandatory except break-glass with incident (§2.4). | +| `RG-06` | Break-glass requires a pre-declared incident and post-hoc audit (§2.5). | +| `RG-07` | Unilateral LLM/operator full restart with active peers is forbidden (§2.6). | +| `RG-08` | Ambiguous policy state denies restart (§4). | + +The `restart-governance/v1` **policy version** field is emitted on future +restart audit events so approvals can be reconciled against the policy revision +in force. + +## 6. Dogfooding + +Gitea-Tools governs its own MCP control plane by this policy. Author, reviewer, +merger, and reconciler sessions operating on this repository use the recovery +ladder (§2.2) — reconnect and rebind, never self-restart — and any real restart +of the Gitea-Tools stable control runtime follows the controller-approval + +drain path defined here. + +## 7. Acceptance and cross-links + +This ADR is the authoritative restart-governance policy. It **must** stay +cross-linked from the safety model and the web-console deployment boundary: + +- `docs/safety-model.md` § Process restart governance references this ADR. +- `docs/webui-deployment.md` references this ADR for restart/reload disposition. + +It is linked to its issue lineage — umbrella **#655**, vision **#652**, roadmap +**#653**, contamination guard **#630**, and console restart UX **#642** — in +§ Related above. + +## 8. Non-goals + +- Implementing the restart coordinator or approval state machine (#630, later + children of #655). +- Implementing HA multi-instance restart or quorum machinery. +- Introducing any process-kill or auto-restart tool; existing auto-restart + behavior must be inventoried before any new restart tool is enabled. diff --git a/docs/mcp-restart-coordinator.md b/docs/mcp-restart-coordinator.md new file mode 100644 index 0000000..b1368c0 --- /dev/null +++ b/docs/mcp-restart-coordinator.md @@ -0,0 +1,95 @@ +# MCP restart coordinator and impact analysis (#658) + +Before any sanctioned MCP restart, a central coordinator evaluates the live +control-plane state and produces an **impact preview** so operators and the web +console (#642 / #652) can see the blast radius *before* concurrent LLM work is +disrupted. Uncoordinated restarts destroy in-flight author/reviewer/merger work +and give operators no way to see what they are about to break. + +This lands the coordinator + impact DTO + a dry-run MCP tool. It is the single +sanctioned entry point for restart evaluation post-#657 (which inventoried the +restart/reload/kill paths). The **mutative apply** path — actually performing a +restart — is a later child gated by a drain proof and is explicitly out of +scope here. + +## Components + +| Piece | Where | Responsibility | +|-------|-------|----------------| +| `restart_coordinator.evaluate_restart_impact` | `restart_coordinator.py` | Pure classification: inventory → impact report DTO. No I/O, no restart. | +| `RestartImpactReport` / `SessionImpact` / `LeaseImpact` | `restart_coordinator.py` | Console-facing DTO (`.as_dict()` is JSON-serializable). | +| `ControlPlaneDB.list_sessions` | `control_plane_db.py` | Read-only session inventory (the process-level unit a restart kills). | +| `gitea_request_mcp_restart` | `gitea_mcp_server.py` | MCP tool: gathers inventory from the #613 DB, calls the coordinator, returns the report. Dry-run only. | + +## Dimensions evaluated + +The coordinator classifies the inventory across the dimensions #658 requires: + +- **Sessions** — every active MCP session; a restart terminates all of them. + Liveness = `status == active` **and** the owner pid is alive **and** the + heartbeat is fresh (default window 15 min). Dead/stale sessions do not count + toward blast radius. +- **Leases / locks** — control-plane leases joined with work items and their + freshness (`lease_lifecycle.classify_lease_freshness`). Only `active` (live + owner) leases are *disruptive*; expired / released / dead-process leases never + withhold a restart. +- **Issue / PR work** — the issues and PRs behind disruptive leases. +- **Mutations / critical sections** — a live lease carrying an author worktree + or a mutating phase (`implementing`, `publishing`, `merging`, …) is a + critical section a restart must not sever. +- **Terminal (merge) lock** — an active terminal lock always makes a restart + unsafe. +- **Prior recovery attempts** — narrower recovery already tried (e.g. sanctioned + client reconnects) is echoed so the operator sees the escalation history. + +## Verdict + +Exactly three verdicts, matching the acceptance criteria: + +| Verdict | `allow_restart` | Meaning | +|---------|-----------------|---------| +| `safe` | `true` | No other live sessions, no live leases, no terminal lock. | +| `unsafe` | `false` | Live work would be disrupted and no operator override is present — **or** the inventory could not be completed (fail closed). | +| `override` | `true` | Live work present, but an operator override accepts the blast radius. | + +`override_would_allow` tells the console whether an override path exists for the +current state. `blast_radius` is a `none` / `low` / `medium` / `high` severity +band derived from the affected session and work counts. + +### Fail closed + +If the control-plane inventory cannot be completed (DB unavailable, a listing +failed), `inventory_complete` is `false` and the verdict is `unsafe` / deny. An +incomplete evaluation must never green-light a restart. + +### Operator override authority + +Override authority is read from the environment variable +`GITEA_OPERATOR_RESTART_OVERRIDE_AUTHORIZATION` and **never** from a tool +argument. A worker session cannot set an environment variable on an +already-running daemon, so override cannot be self-asserted (same pattern as the +#630 daemon-maintenance authorization). The `request_override` tool argument only +expresses caller intent; it takes effect solely when the environment +authorization is present. + +## The tool + +```text +gitea_request_mcp_restart(remote, host, org, repo, + dry_run=True, request_override=False, + session_id=None, limit=200) +``` + +Read-only, dry-run, and it **never restarts anything**. `apply_supported` is +always `false`; passing `dry_run=False` performs no restart and reports that +apply is gated by a drain proof (a separate child). + +## Audit + +Every evaluation carries an `audit_record` (event, coordinator version, verdict, +allow decision, blast radius, counts, timestamp) so restart decisions are +auditable. No secrets flow through the coordinator — session ids, pids, and +profiles are operational metadata only. + +A representative dry-run report is in +[`mcp-restart-impact-sample.json`](./mcp-restart-impact-sample.json). diff --git a/docs/mcp-restart-impact-sample.json b/docs/mcp-restart-impact-sample.json new file mode 100644 index 0000000..e0136ee --- /dev/null +++ b/docs/mcp-restart-impact-sample.json @@ -0,0 +1,148 @@ +{ + "coordinator_version": "1.0.0-issue-658", + "evaluated_at": "2026-07-24T06:00:00+00:00", + "dry_run": true, + "restart_performed": false, + "inventory_complete": true, + "incomplete_reasons": [], + "verdict": "unsafe", + "allow_restart": false, + "override_would_allow": true, + "operator_override": false, + "blast_radius": "high", + "reasons": [ + "live work would be disrupted; restart denied without operator override", + "1 critical section(s) in flight (active lease with a live owner)" + ], + "affected_sessions": [ + { + "session_id": "prgs-author-30988-d6f43c25", + "role": "author", + "profile": "prgs-author", + "pid": 1, + "status": "active", + "alive": true, + "heartbeat_stale": false, + "is_requester": false, + "live": true + }, + { + "session_id": "prgs-reviewer-4157-0ce9", + "role": "reviewer", + "profile": "prgs-reviewer", + "pid": 1, + "status": "active", + "alive": true, + "heartbeat_stale": false, + "is_requester": true, + "live": true + } + ], + "affected_leases": [ + { + "lease_id": "lease-abc", + "session_id": "prgs-author-30988-d6f43c25", + "role": "author", + "phase": "implementing", + "freshness": "active", + "work_kind": "issue", + "work_number": 658, + "worktree_path": "/repo/branches/feat-issue-658", + "disruptive": true, + "is_mutation": true, + "is_critical_section": true + }, + { + "lease_id": "lease-dead", + "session_id": "prgs-author-91485", + "role": "author", + "phase": "allocated", + "freshness": "stale_dead_process", + "work_kind": "issue", + "work_number": 651, + "worktree_path": null, + "disruptive": false, + "is_mutation": false, + "is_critical_section": false + } + ], + "critical_sections": [ + { + "lease_id": "lease-abc", + "session_id": "prgs-author-30988-d6f43c25", + "role": "author", + "phase": "implementing", + "freshness": "active", + "work_kind": "issue", + "work_number": 658, + "worktree_path": "/repo/branches/feat-issue-658", + "disruptive": true, + "is_mutation": true, + "is_critical_section": true + } + ], + "affected_issues": [ + 658 + ], + "affected_prs": [], + "mutations": [ + { + "lease_id": "lease-abc", + "session_id": "prgs-author-30988-d6f43c25", + "role": "author", + "phase": "implementing", + "freshness": "active", + "work_kind": "issue", + "work_number": 658, + "worktree_path": "/repo/branches/feat-issue-658", + "disruptive": true, + "is_mutation": true, + "is_critical_section": true + } + ], + "terminal_lock": null, + "ack_state": { + "prgs-author-30988-d6f43c25": "pending" + }, + "prior_recovery_attempts": [ + { + "kind": "client_reconnect", + "at": "2026-07-24T06:00:00+00:00", + "outcome": "insufficient" + } + ], + "counts": { + "sessions_total": 2, + "sessions_live_other": 1, + "leases_total": 2, + "leases_disruptive": 1, + "critical_sections": 1, + "mutations": 1, + "affected_issues": 1, + "affected_prs": 0, + "prior_recovery_attempts": 1 + }, + "audit_record": { + "event": "restart_impact_evaluated", + "coordinator_version": "1.0.0-issue-658", + "evaluated_at": "2026-07-24T06:00:00+00:00", + "dry_run": true, + "operator_override": false, + "requesting_session_id": "prgs-reviewer-4157-0ce9", + "inventory_complete": true, + "verdict": "unsafe", + "allow_restart": false, + "blast_radius": "high", + "counts": { + "sessions_total": 2, + "sessions_live_other": 1, + "leases_total": 2, + "leases_disruptive": 1, + "critical_sections": 1, + "mutations": 1, + "affected_issues": 1, + "affected_prs": 0, + "prior_recovery_attempts": 1 + } + } +} diff --git a/docs/mcp-restart-path-inventory.md b/docs/mcp-restart-path-inventory.md new file mode 100644 index 0000000..c880269 --- /dev/null +++ b/docs/mcp-restart-path-inventory.md @@ -0,0 +1,90 @@ +# MCP restart / reload / kill path inventory (#657) + +Complete inventory of every code, script, and host path that can **restart, +reload, reconnect, kill, or force-recreate** an MCP process in this project, +with each path classified and linked to the guard that constrains it. + +This document is the human-readable companion to the machine-readable registry +in [`mcp_restart_paths.py`](../mcp_restart_paths.py). The two are kept in +lock-step by [`tests/test_mcp_restart_paths.py`](../tests/test_mcp_restart_paths.py): +every `path_id` below must appear in this file, and the source guards are run +against the live tree. + +Roadmap linkage: this inventory is the enumeration step of the restart +governance work — parent **#655**, restart-governance ADR **#656**, vision +**#652**, roadmap **#653**. Related detection/guard work: master-advance +staleness **#591**/**#420**, side-effect-free resolver **#685**, transport flap +**#584**, manual-kill contamination **#630**. + +## Classifications + +| Classification | Meaning | +|---|---| +| `sanctioned_narrow_recovery` | One-shot, safe-by-construction recovery that never targets the running daemon. | +| `guarded_fail_closed` | Detects a restart-requiring condition, then fails mutations closed and emits reconnect guidance. Never self-restarts. | +| `forbidden` | A workflow-safety violation; where an LLM tool could invoke it, it is marked contamination. | +| `removed` | A previously-existing unguarded restart primitive that has been deleted; a regression guard keeps it absent. | +| `host_residual` | Behavior owned by the host/IDE, outside this process's control. Documented, not code-guarded here. | + +## The rule + +**No component may perform an unguarded full restart of the MCP daemon.** The +in-process daemon (`gitea_mcp_server.py`, `mcp_server.py`, +`role_session_router.py`) must never replace or terminate its own process: +replacing the process after the host has wired up the stdio pipes desyncs the +JSON-RPC transport (observed with Antigravity/Cascade hosts). Recovery is owned +by the host/operator via a client reconnect — the daemon only ever *detects* +and *fails closed*. + +## Inventory + +| path_id | Classification | Mechanism | Guard | Refs | +|---|---|---|---|---| +| `cli_venv_bootstrap_execv` | sanctioned_narrow_recovery | CLI wrapper scripts re-exec into `venv/bin/python3` via `os.execv`, guarded by `sys.executable != venv_python`. | One-shot pre-import bootstrap; runs before any MCP transport exists and only when not already on the venv interpreter; idempotent guard prevents a re-exec loop. | #657 | +| `daemon_self_replacement` | forbidden | The daemon replacing/terminating its own process (`os.execv`/`os.kill`/`os._exit`) to reload code. | Forbidden by design; enforced against the source tree by `assert_no_daemon_self_replacement()`. | #657, #584 | +| `legacy_auto_restart_helper` | removed | A helper (`_trigger_mcp_auto_restart`) that actively restarted the server from the read-only resolver path. | Removed in #685; kept absent by `assert_auto_restart_helper_absent()`. | #685, #657 | +| `config_touch_reload` | removed | Touching (utime) the MCP client config to make the host reload the server. | Removed from the resolver in #685: stale detection is report-only, never mutating config, spawning threads, or calling `os._exit`. | #685, #657 | +| `master_advance_auto_restart` | guarded_fail_closed | On-disk master advancing past the running code. | `master_parity_gate` captures startup parity and blocks mutations while stale, emitting restart guidance; the process never self-restarts. | #420, #591, #657 | +| `stale_runtime_resolver_reconnect` | guarded_fail_closed | The capability resolver detecting a stale serving process. | Report-only (#685): returns `restart_required`/`stop_required` and an exact reconnect action; no restart, thread, config touch, or `os._exit`. | #685, #657 | +| `manual_daemon_kill` | forbidden | Shell kills of the daemon: `pkill -f mcp_server.py`, `killall`, broad `pkill -f python` sweeps, or `kill ` of a daemon pid. | Forbidden (#630): `runtime_recovery_guard` classifies these as contamination and `gitea_record_daemon_process_kill_attempt` writes a durable marker that fails later mutations closed. Operator maintenance authorization is read only from the environment. | #630, #657 | +| `conflict_marker_infra_stop` | guarded_fail_closed | The daemon entrypoint scans for unresolved merge-conflict markers at startup and stops (`sys.exit(1)`). | Fail-closed startup stop, not a restart: the process exits and waits for the operator to resolve conflicts and relaunch; never loops. | #657 | +| `ide_client_reconnect` | host_residual | A manual `/mcp reconnect` (or equivalent host action) that recreates the MCP client connection. | Outside this process's control; the sanctioned recovery the gates point operators toward. No in-process code initiates it. | #584, #656, #657 | +| `profile_switch_runtime` | sanctioned_narrow_recovery | Switching the active execution profile at runtime (dynamic-profile mode). | In-process and restart-free: `runtime_switching_supported` is true, so a switch rebinds capability without recreating the process. | #656, #657 | + +## Guards enforced in CI + +`tests/test_mcp_restart_paths.py` asserts, against the live source tree: + +1. **Registry well-formedness** — every path has a valid classification, a + non-empty guard description, references, and locations; ids are unique; all + five classifications are represented. +2. **Unknown restart attempts fail closed** — + `assert_restart_attempt_registered()` raises `UnknownRestartPathError` for + any path id not in this inventory, so a novel/unnamed restart primitive + cannot slip through silently. +3. **Daemon never self-replaces** — `assert_no_daemon_self_replacement()` scans + the daemon modules for `os.execv`/`os.kill`/`os._exit`/`os.abort` calls + (comment/docstring mentions are ignored) and finds none. +4. **Legacy helper stays removed** — `assert_auto_restart_helper_absent()` + confirms `_trigger_mcp_auto_restart` has not returned. +5. **pkill stays forbidden** — a daemon `pkill` command still classifies as + contamination via `runtime_recovery_guard`. + +## Residual host behaviors (outside process control) + +* `/mcp reconnect` in the IDE/host — the sanctioned recovery for stale-runtime, + transport-flap (#584), and worktree-binding conditions. The daemon can only + emit guidance toward it. +* Host-level process management (the operator relaunching the daemon after a + fail-closed stop, or after resolving merge conflicts). + +These are documented rather than code-guarded because the process cannot +observe or gate them from inside itself. + +## Rollout + +Per #657, guards are introduced flag-free as **regression assertions** (they +codify invariants that already hold) before any hard runtime block is layered +on. When the restart coordinator (#655/#656) lands, registered paths gain a +coordinator token/capability check; unregistered attempts already fail closed +today via `assert_restart_attempt_registered()`. diff --git a/docs/mcp-tool-inventory.md b/docs/mcp-tool-inventory.md index 8cb865b..b1aed12 100644 --- a/docs/mcp-tool-inventory.md +++ b/docs/mcp-tool-inventory.md @@ -135,6 +135,7 @@ that gates each call, not which tools exist. - `gitea_release_merger_pr_lease` - `gitea_release_reviewer_pr_lease` - `gitea_release_workflow_lease` +- `gitea_request_mcp_restart` - `gitea_resolve_task_capability` - `gitea_resume_review_draft` - `gitea_review_pr` diff --git a/docs/safety-model.md b/docs/safety-model.md index 31c740a..bbab241 100644 --- a/docs/safety-model.md +++ b/docs/safety-model.md @@ -46,3 +46,17 @@ If shell helpers are unavailable and MCP commit cannot run, stop with a recovery report (restart session, clear hung terminals, use MCP-native commit). See [`llm-workflow-runbooks.md`](llm-workflow-runbooks.md) § MCP-native commit path (#260) and agent temp artifact cleanup (#261). + +## 7. Process restart governance + +Restarting the MCP control-plane process is destructive to concurrent multi-role +work and is governed by a dedicated policy. Restart is a **last resort** behind +narrower recoveries (reconnect, rebind), full/host restart is reserved to +operator/admin under **controller approval + automated safety gates**, a +unilateral LLM or operator full restart with active peers is **forbidden**, and +ambiguous policy state **denies** restart. Break-glass is a separate, +incident-backed path with a mandatory audit. + +See [`architecture/mcp-restart-governance.md`](architecture/mcp-restart-governance.md) +(#656) for the authorization matrix, the recovery ladder, break-glass +conditions, and the `RG-01`–`RG-08` policy IDs. diff --git a/docs/webui-deployment.md b/docs/webui-deployment.md index fa754ad..2ad46ab 100644 --- a/docs/webui-deployment.md +++ b/docs/webui-deployment.md @@ -55,6 +55,15 @@ shipped to the browser. assumption paths, and the client-secret policy. Use it to verify an instance is configured for internal-only operation. +## Process restart / reload disposition + +The console never exposes a restart or reload control; process restart of the +MCP control-plane runtime is governed separately. Restart is a last resort behind +reconnect/rebind, full restart is operator/admin-only under controller approval +plus safety gates, and break-glass is an incident-backed path. See +[`architecture/mcp-restart-governance.md`](architecture/mcp-restart-governance.md) +(#656). + ## Non-goals (MVP) - Full SSO or session login in the UI diff --git a/docs/webui-local-dev.md b/docs/webui-local-dev.md index 34ea5ce..cad3507 100644 --- a/docs/webui-local-dev.md +++ b/docs/webui-local-dev.md @@ -54,6 +54,7 @@ status, onboarding checklist state, and the fail-closed error payloads (#635). | `/` | Home / operator overview | | `/health` | JSON liveness (`status`, `service`, `mode`, `timestamp`, `uptime_seconds`) | | `/api/v1/system/health` | Structured read-only system health (#634) | +| `/system-health` | System-health dashboard — readiness, version/uptime, dependencies, MCP namespaces, stale-runtime parity (#639) | | `/queue` | Live PR and issue queue dashboard (#429) | | `/api/queue` | JSON queue export with pagination metadata | | `/projects` | Project registry list with status and onboarding progress (#427, #635) | @@ -76,6 +77,11 @@ status, onboarding checklist state, and the fail-closed error payloads (#635). | `/api/actions/{id}/preview` | Mutation ledger preview (GET, read-only) | | `/leases` | Lease and collision visibility (#433) | | `/api/leases` | JSON lease/collision export | +| `/sessions` | Phase 1 shell stub — session inventory (backed by #636) | +| `/inventory` | Phase 1 shell stub — unified inventory (backed by #636) | +| `/timeline` | Phase 1 shell stub — workflow event timeline | +| `/policy` | Phase 1 shell stub — capability/role policy placeholder | +| `/insights` | Phase 1 shell stub — operational insights placeholder | Most routes are GET-only. POST/PUT/PATCH/DELETE return `405` with `read-only-mvp`, except `/audit` and `/api/audit` which accept POST for @@ -248,6 +254,57 @@ through the console redaction pass before it is emitted, so a planted or accidental secret degrades to the placeholder rather than reaching a client. The view never edits policy and exposes no gate-weakening toggle. +## Application shell — Phase 1 (#638) + +The console shell (`webui/layout.py`) renders a grouped navigation driven by a +single nav-config module, `webui/nav.py`. Nav groups follow the epic #631 +Phase 1 information architecture: **Health, Traffic, Runtime/Sessions, +Projects, Inventory, Timeline, Policy** (placeholder), and **Insights** +(placeholder). Live views and Phase 1 placeholders (`stub`) are declared in one +place so the layout and the route table cannot drift. + +The header carries two read-only status badges — an **environment** badge +(`local` for loopback binds, `remote` otherwise, derived from `WEBUI_HOST`) and +a **mode: read-only** badge — plus a **Docs** link to this document. No +privileged action controls are present in the Phase 1 shell. + +Not-yet-implemented surfaces (`/sessions`, `/inventory`, `/timeline`, +`/policy`, `/insights`) resolve to graceful read-only stub pages instead of +404s; their backing views land in later child issues of #631 (the inventory +surfaces are backed by #636). Mutating methods on stub routes still fail closed +with `read-only-mvp`. + +## System-health dashboard (#639) + +`/system-health` renders the same snapshot the `/api/v1/system/health` API +returns, so the page and the API can never disagree. Cards: overall readiness, +stale-runtime parity, version and uptime, dependency probes, MCP namespaces, +probe errors (only when present), and recovery pointers. `?deep=1` opts into +the network probe exactly as the API does; the plain page load stays cheap. + +Field authority and honesty rules: + +* `ready` and `readiness_complete` are shown separately. A snapshot whose + required probes never ran is not the same as one that ran them and passed, + and the page never collapses the two into an unproven green. +* A probe that did not run appears under **Not probed**, never as healthy. +* `stale_runtime.mutation_safe` is displayed verbatim from the API. When the + runtime is stale, or when parity is indeterminate, the page warns and does + not claim mutation safety. +* MCP namespaces are reported `unproven`: the web process runs outside the + IDE-managed MCP client and cannot prove that path (#543). + +Redaction is split by field kind. Free text — probe details, readiness and +parity reasons, probe errors — passes through `system_health.redact`. +Structured fields — commit SHAs, probe names, statuses, timestamps — are +HTML-escaped only, because `redact`'s opaque-token rule matches any run of 32 +or more characters and would otherwise blank every 40-character git SHA, which +is precisely the evidence the parity view exists to show. + +The dashboard is read-only: no restart, reload, or process-kill control. Those +arrive in Phase 2 (#642). Recovery guidance points at the sanctioned client +reconnect / operator restart path — never a manual daemon kill (#630). + ## Deployment boundary (#435) MVP serves on loopback by default. Binding `0.0.0.0` or `::` is **refused** @@ -307,6 +364,108 @@ health, workflow/schema SHA-256 hashes, and stale-runtime warnings when the checkout is behind merged safety-gate changes. Restart guidance links to #420; no tokens or MCP restart actions are exposed. +## Workflow-event timeline (#637) + +`GET /api/v1/timeline` is a read-only, versioned aggregation of workflow +events from every available source into one normalised, filterable stream. It +is the model layer for the Phase 1 timeline console view (a later child issue +of #631); this issue ships the schema, adapters, and read API only. + +### Schema (versioned) + +`webui/timeline.py` declares `TIMELINE_SCHEMA_VERSION` (currently `1`) and the +frozen `WorkflowEvent` record. Every response carries `schema_version` so a +consumer can branch on shape. One event: + +```json +{ + "source": "control_plane", + "event_type": "lease.renew", + "event_key": "cp:1421", + "timestamp": "2026-07-23T02:00:00Z", + "actor": null, + "role": null, + "issue_number": 637, + "pr_number": null, + "session_id": null, + "tool_name": null, + "decision": null, + "message": "lease renewed", + "correlation_id": "issue#637", + "evidence_refs": [], + "sensitive": true +} +``` + +`event_key` is stable and unique per source (`cp:`, +`cth:::`), so pagination and dedup are deterministic. + +### Sources and field authority + +| Source | Adapter | Authority | +|---|---|---| +| Control-plane `events` ⋈ `work_items` | `adapt_cp_events` | `event_type`, `message`, `timestamp`, issue/PR scope come from the CP database, read through a `mode=ro` URI (never creates the DB or runs migrations) | +| Gitea Canonical Thread Handoff comments | `adapt_cth_comments` | `actor`, `role` (next owner), `decision`, `evidence_refs`, `timestamp` come from the parsed CTH comment body (`canonical_thread_handoff`) | + +Handoff comments are thread-scoped: they are only read when the request filters +by a single `issue` or `pr`. Otherwise the handoff source reports `not run` +with a reason — it is never rendered as empty-and-healthy. Each source degrades +independently: an unavailable control-plane DB or a failed comment fetch is a +`sources[]` entry with `ok:false` and a `reason`, never a dropped timeline. + +### Query parameters + +`issue`, `pr`, `session` (conjunctive filters); `limit` (default 50, max 500) +and `offset` for pagination; `remote`, `org`, `repo` to override the default +registry-project scope. Events sort ascending by +`(timestamp, source_rank, event_key)`; missing timestamps sort last. + +### Filter authority, and refusing what cannot be answered + +A filter dimension is only meaningful for a source whose records carry it. +Each source declares its own support in `_SOURCE_FILTER_SUPPORT` and reports it +per response as `supported_filters` / `unsupported_filters`: + +| Source | issue | pr | session | +|---|---|---|---| +| `control_plane` | yes | yes | **no** — the `events` table is `(event_id, work_item_id, event_type, message, created_at)` and records no session | +| `gitea_handoff` | yes | yes | yes — a CTH comment declares its own `Session:` field | + +`session_id` is read only from that declared CTH field. It is never inferred +from a work item, an actor, or message text, and a value that is +redaction-altering or bare-secret-shaped is dropped rather than emitted. + +When **no source that ran** can carry a requested dimension, the request is +refused rather than answered: the response is `422` with `ok:false` and a +structured `error` naming `unsupported_filters` and the per-source reason. A +`200` with zero events would tell an operator that no such activity exists, +which is a stronger — and false — claim than "this cannot be answered here". +A source that *can* answer the dimension and simply matched nothing still +returns `200` with `ok:true` and an empty page. + +### Redaction + +Every free-text field (event messages, decision/proof text, roles, actors) is +passed through the console redaction policy (`webui.console_redaction`, backed +by `gitea_audit.redact`) before it leaves the module, failing closed to the +placeholder. No unredacted tool arguments or secrets are ever emitted, and a +generation error never drops raw data to a caller or a log. + +Redaction also runs *before* any structured value is derived from free text. +`evidence_refs` are extracted from already-redacted proof/decision text, and a +commit reference is recognised only where the text declares one (`commit`, +`head`, `base`, `sha`, …). An undeclared 40-character hex run has the exact +shape of a Gitea access token, so it is never lifted out of prose into a +structured field. Every reference is then independently revalidated against an +allowed shape and a second redaction pass immediately before serialization; +anything unproven is dropped and the event is flagged `sensitive`. + +### Tests + +```bash +pytest tests/test_webui_timeline.py -q +``` + ## Tests ```bash diff --git a/gitea_mcp_server.py b/gitea_mcp_server.py index 87360bf..0199910 100644 --- a/gitea_mcp_server.py +++ b/gitea_mcp_server.py @@ -440,13 +440,32 @@ def _session_author_lock_worktree() -> str | None: Used to derive the author mutation workspace when no explicit ``worktree_path`` or env binding is provided. Never invents a path. + + #864: a session pointer whose owner PID is dead and is not this process + must not force workspace binding for other issues — rebind is required for + that issue, and a stale dead-owner pointer must not poison unrelated work. """ try: lock = issue_lock_store.read_session_issue_lock() or {} except Exception: return None path = (lock.get("worktree_path") or "").strip() - return path or None + if not path: + return None + pid = lock.get("session_pid") + if pid is None: + pid = lock.get("pid") + try: + pid_i = int(pid) if pid is not None else None + except (TypeError, ValueError): + pid_i = None + if ( + pid_i is not None + and pid_i != os.getpid() + and not issue_lock_store.is_process_alive(pid_i) + ): + return None + return path def _resolve_preflight_workspace_path(worktree_path: str | None = None) -> str: @@ -1450,7 +1469,7 @@ def verify_preflight_purity( dirty_files = sorted( _parse_porcelain_entries(_get_workspace_porcelain(workspace)) ) - if dirty_files: + if dirty_files and task != "commit_files": raise RuntimeError( nwb.format_namespace_workspace_binding_error( role_kind=role, @@ -2021,6 +2040,7 @@ import dependency_graph # noqa: E402 # #784 durable dependency edges import control_plane_db # noqa: E402 import lease_lifecycle # noqa: E402 import workflow_dashboard # noqa: E402 # #605 live queue/lease dashboard +import restart_coordinator # noqa: E402 # #658 MCP restart coordinator/impact import incident_bridge # noqa: E402 import sentry_observability # noqa: E402 (#606 optional Sentry observability) import sentry_incident_bridge # noqa: E402 (#607 Sentry→Gitea incident bridge) @@ -2031,6 +2051,8 @@ import issue_lock_store # noqa: E402 import issue_lock_adoption # noqa: E402 import issue_lock_recovery # noqa: E402 import issue_lock_renewal # noqa: E402 +import dirty_orphan_worktree_recovery # noqa: E402 # #860 dirty orphan recovery +import dirty_same_claimant_session_rebind # noqa: E402 # #864 import stacked_pr_support # noqa: E402 import merge_approval_gate # noqa: E402 import review_quarantine # noqa: E402 # #695 contaminated formal-review quarantine @@ -2425,6 +2447,20 @@ def _evaluate_issue_lock_recovery( descendant_sha=local_head, ) + # #871: the inverse of the #768 descendant relation — the *remote* head may + # have advanced past the local/recorded head via a sanctioned merge-based + # branch sync (``gitea_update_pr_branch_by_merge``) while the local worktree + # stayed put. Observe that provenance server-side so the assessor can prove + # it and nothing else. Probed only when the heads differ; never from any + # caller-supplied value. + sync_provenance: dict | None = None + if remote_head and local_head and remote_head != local_head: + sync_provenance = issue_lock_worktree.read_merge_sync_provenance( + worktree_path, + prior_head_sha=local_head, + synced_head_sha=remote_head, + ) + # #772: with no remote branch there is no head to measure against, so the # base the branch was cut from is observed instead. Probed only in that # case, so the published path's evidence is untouched (#772 AC8). @@ -2467,6 +2503,7 @@ def _evaluate_issue_lock_recovery( remote_branch_exists=remote_branch_exists, recorded_base_sha=recorded_base, base_ancestry=base_ancestry, + sync_provenance=sync_provenance, ) @@ -4342,6 +4379,539 @@ def gitea_lock_issue( return result +@mcp.tool() +@mcp.tool() +def gitea_recover_dirty_orphaned_issue_worktree( + issue_number: int, + branch_name: str, + source_worktree_path: str, + expected_local_head: str, + expected_remote_head: str, + expected_dirty_fingerprints: dict, + remote: str = "dadeschools", + host: str | None = None, + org: str | None = None, + repo: str | None = None, + recovery_worktree_path: str | None = None, + dry_run: bool = False, +) -> dict: + """Recover a dirty orphaned same-claimant author issue worktree (#860). + + Explicit recovery operation — does **not** silently widen ``gitea_lock_issue``. + + Accepts authoritative expected pins (repository, issue, branch, source + worktree, claimant, local head, remote/PR head, dirty fingerprints) and + fails closed on any mismatch. PID-less malformed locks are never treated + as live merely because expiry is absent. The source worktree is frozen; + recovery prepares a separate worktree at the pinned remote head, re-applies + dirty bytes with path-level conflict detection, and binds a live author + session only after recovery state is consistent. + + Args: + issue_number: Issue whose durable claim is being recovered. + branch_name: Locked branch ``(fix|feat|docs|chore)/issue-N-…``. + source_worktree_path: Registered dirty source worktree under branches/. + expected_local_head: Full 40-char SHA of the source worktree HEAD. + expected_remote_head: Full 40-char SHA of the remote/PR head to sync to. + expected_dirty_fingerprints: ``{relative_path: sha256}`` of dirty bytes. + remote/host/org/repo: Repository binding. + recovery_worktree_path: Optional recovery worktree path under branches/. + dry_run: Assess eligibility only; no filesystem or lock mutation. + + Returns: + dict with success, outcome, conflicts, recovery_worktree_path, reasons, + evidence, and journal metadata. + """ + task = "recover_dirty_orphaned_issue_worktree" + ok, block_reasons = role_session_router.check_author_mutation_after_reviewer_stop( + task + ) + if not ok: + return { + "success": False, + "performed": False, + "outcome": "REFUSED", + "reasons": block_reasons, + } + blocked = _namespace_mutation_block(task, remote=remote) + if blocked: + return blocked + blocked = _profile_permission_block( + task_capability_map.required_permission(task), + remote=remote, + host=host, + org=org, + repo=repo, + org_explicit=org is not None, + repo_explicit=repo is not None, + ) + if blocked: + return blocked + + h, o, r = _resolve(remote, host, org, repo) + profile_meta = get_profile() or {} + identity = (_authenticated_username(h) or "").strip() + profile = (profile_meta.get("profile_name") or "").strip() + if not identity or not profile: + return { + "success": False, + "performed": False, + "outcome": "REFUSED", + "reasons": ["could not resolve authenticated identity/profile"], + } + + existing_lock = _load_existing_issue_lock( + remote=remote, org=o, repo=r, issue_number=issue_number + ) + + src = os.path.realpath(source_worktree_path) + git_state = issue_lock_worktree.read_worktree_git_state(src) + observed_local = (git_state.get("head_sha") or "").strip() + porcelain = git_state.get("porcelain_status") or "" + current_branch = git_state.get("current_branch") + + # Observed dirty fingerprints from source worktree bytes. + observed_fps: dict[str, str] = {} + dirty_contents: dict[str, bytes] = {} + for rel in (expected_dirty_fingerprints or {}): + rel_n = str(rel).strip() + fpath = os.path.join(src, rel_n) + if not os.path.isfile(fpath): + continue + with open(fpath, "rb") as fh: + data = fh.read() + dirty_contents[rel_n] = data + observed_fps[rel_n] = dirty_orphan_worktree_recovery.sha256_bytes(data) + + # Remote head observation (best-effort; pin mismatch fails closed). + observed_remote = "" + try: + probe = subprocess.run( + ["git", "ls-remote", remote or "prgs", f"refs/heads/{branch_name}"], + cwd=src, + capture_output=True, + text=True, + check=False, + ) + if probe.returncode == 0 and (probe.stdout or "").strip(): + observed_remote = (probe.stdout or "").strip().split()[0] + except Exception: + observed_remote = "" + + registered = False + try: + listing = subprocess.run( + ["git", "worktree", "list", "--porcelain"], + cwd=src, + capture_output=True, + text=True, + check=False, + ) + if listing.returncode == 0: + registered = src in (listing.stdout or "") + except Exception: + registered = False + + project_root = _canonical_local_git_root() + canonical_root = author_mutation_worktree.resolve_canonical_repo_root( + src, project_root + ) + + competing_locks: list[dict] = [] + try: + all_live = issue_lock_store.list_live_locks() + for l in all_live: + if l.get("issue_number") == issue_number: + wt = l.get("worktree_path") + if not wt or not issue_lock_store._same_realpath(wt, src): + competing_locks.append(l) + except Exception: + competing_locks = [] + + wf_active = False + wf_expired = True + try: + db, _ = _control_plane_db_or_error() + if db is not None: + active_leases_data = lease_lifecycle.list_active_leases( + db, + remote=remote if remote in REMOTES else remote, + org=o, + repo=r, + ) + leases_list = active_leases_data.get("leases") or [] + for l in leases_list: + if l.get("work_number") == issue_number and l.get("work_kind") == "issue": + fresh = l.get("freshness") or {} + if fresh.get("status") == "active": + wf_active = True + wf_expired = False + elif fresh.get("status") in ("expired", "stale_dead_process"): + wf_active = False + wf_expired = True + except Exception: + pass + + assessment = dirty_orphan_worktree_recovery.assess_dirty_orphan_recovery( + existing_lock, + issue_number=issue_number, + branch_name=branch_name, + source_worktree_path=src, + remote=remote if remote else "prgs", + org=o, + repo=r, + identity=identity, + profile=profile, + expected_local_head=expected_local_head, + expected_remote_head=expected_remote_head, + expected_dirty_fingerprints=expected_dirty_fingerprints or {}, + current_branch=current_branch, + porcelain_status=porcelain, + observed_local_head=observed_local, + observed_remote_head=observed_remote, + observed_dirty_fingerprints=observed_fps, + competing_live_locks=competing_locks, + competing_live_sessions=[], + workflow_lease_active=wf_active, + workflow_lease_expired=wf_expired, + canonical_repo_root=canonical_root, + worktree_registered=registered, + current_pid=os.getpid(), + ) + if dry_run or not assessment.get("eligible"): + return { + "success": bool(assessment.get("eligible")), + "performed": False, + "dry_run": dry_run, + "outcome": assessment.get("outcome"), + "reasons": list(assessment.get("reasons") or []), + "evidence": dict(assessment.get("evidence") or {}), + "eligible": bool(assessment.get("eligible")), + } + + if not recovery_worktree_path: + recovery_worktree_path = os.path.join( + canonical_root, + "branches", + f"recovery-issue-{issue_number}-dirty-orphan", + ) + + # Load blob contents at local/remote heads for conflict detection. + def _blob_at(head: str, rel: str) -> bytes | None: + try: + proc = subprocess.run( + ["git", "show", f"{head}:{rel}"], + cwd=src, + capture_output=True, + check=False, + ) + if proc.returncode != 0: + return None + return proc.stdout + except Exception: + return None + + local_contents = { + rel: _blob_at(expected_local_head, rel) + for rel in (expected_dirty_fingerprints or {}) + } + remote_contents = { + rel: _blob_at(expected_remote_head, rel) + for rel in (expected_dirty_fingerprints or {}) + } + + # Preflight purity is satisfied via explicit worktree_path on this tool's + # recovery path; source remains frozen and is never cleaned. + result = dirty_orphan_worktree_recovery.run_dirty_orphan_recovery( + assessment=assessment, + existing_lock=existing_lock or {}, + issue_number=issue_number, + branch_name=branch_name, + source_worktree_path=src, + recovery_worktree_path=recovery_worktree_path, + remote=remote if remote else "prgs", + org=o, + repo=r, + identity=identity, + profile=profile, + expected_local_head=expected_local_head, + expected_remote_head=expected_remote_head, + expected_dirty_fingerprints=expected_dirty_fingerprints or {}, + dirty_contents=dirty_contents, + local_head_contents=local_contents, + remote_head_contents=remote_contents, + canonical_repo_root=canonical_root, + bind_lock=True, + session_pid=os.getpid(), + ) + # Surface preflight recognition for recovered provenance. + if result.get("success") and result.get("lock_record"): + result["preflight_provenance"] = ( + dirty_orphan_worktree_recovery.preflight_recognizes_recovered_provenance( + result["lock_record"] + ) + ) + return result + +@mcp.tool() +def gitea_rebind_dirty_same_claimant_author_session( + issue_number: int, + branch_name: str, + worktree_path: str, + old_pid: int, + expected_local_head: str, + expected_remote_head: str, + expected_dirty_paths: list[str], + expected_fingerprints: dict, + remote: str = "dadeschools", + host: str | None = None, + org: str | None = None, + repo: str | None = None, + dry_run: bool = False, + authorize_reconciler_execute: bool = False, +) -> dict: + """Rebind a dirty registered issue worktree to this session (#864). + + Sanctioned only when every pin agrees: same claimant, dead old_pid matching + the durable lock, matching local/remote heads, exact dirty path set, and + per-path sha256 fingerprints. Preserves every tracked/untracked byte. + Does not sync remote, create recovery worktrees, clean, reset, or move heads. + + Role gate: + * author — must match the lock claimant identity/profile + * reconciler — execute only when ``authorize_reconciler_execute=True`` + * reviewer/merger — always refuse + + ``gitea.issue.comment`` (author map entry) is required for mutation; dry_run + still assesses fully but writes nothing. Permission alone is never ownership + proof — every pin is re-checked server-side. + + Args: + issue_number: Tracking issue number on the durable lock. + branch_name: Exact locked branch name. + worktree_path: Registered dirty worktree path (must be under branches/). + old_pid: Dead owner PID recorded on the lock (must match session_pid/pid). + expected_local_head: Full local HEAD sha the caller observed. + expected_remote_head: Full remote-tracking HEAD sha the caller observed. + expected_dirty_paths: Exact set of dirty relative paths (tracked+untracked). + expected_fingerprints: Map of relative path -> sha256 hex of file bytes. + remote: Known instance — 'dadeschools' or 'prgs'. + host/org/repo: Optional target overrides (validated against binding). + dry_run: When true, assess only (no lock/session writes). + authorize_reconciler_execute: Reconciler-only execute gate. + """ + role = _profile_role_kind(get_profile()) + role_norm = (role or "").strip().lower() + + # Permission: authors need comment; dry_run assess is reachable under read + # for diagnosis, but execute always needs comment. Reconciler execute also + # needs comment when authorized. + if dry_run: + read_block = _profile_operation_gate("gitea.read") + if read_block: + return { + "success": False, + "dry_run": True, + "reasons": read_block, + "permission_report": _permission_block_report("gitea.read"), + } + else: + blocked = _profile_permission_block( + task_capability_map.required_permission( + "rebind_dirty_same_claimant_author_session" + ), + issue_number=issue_number, + remote=remote, + host=host, + org=org, + repo=repo, + org_explicit=org is not None, + repo_explicit=repo is not None, + ) + if blocked: + return blocked + + if role_norm in {"reviewer", "merger"}: + return { + "success": False, + "dry_run": bool(dry_run), + "outcome": dirty_same_claimant_session_rebind.REFUSED, + "reasons": [ + f"role '{role_norm}' cannot rebind dirty same-claimant author " + "sessions (fail closed)" + ], + } + if role_norm == "reconciler" and not authorize_reconciler_execute and not dry_run: + return { + "success": False, + "dry_run": False, + "outcome": dirty_same_claimant_session_rebind.REFUSED, + "reasons": [ + "reconciler role requires authorize_reconciler_execute=True " + "to execute dirty same-claimant rebind (fail closed)" + ], + } + + h, o, r = _resolve(remote, host, org, repo) + try: + identity = _authenticated_username(h) + except Exception: + identity = None + profile = get_profile() + profile_name = profile.get("profile_name") + + existing = _load_existing_issue_lock( + remote=remote, org=o, repo=r, issue_number=issue_number + ) + resolved_wt = os.path.realpath(os.path.abspath((worktree_path or "").strip())) + inv = dirty_same_claimant_session_rebind.collect_dirty_inventory(resolved_wt) + + branch_res = subprocess.run( + ["git", "-C", resolved_wt, "branch", "--show-current"], + capture_output=True, + text=True, + check=False, + ) + current_branch = (branch_res.stdout or "").strip() or None + head_res = subprocess.run( + ["git", "-C", resolved_wt, "rev-parse", "HEAD"], + capture_output=True, + text=True, + check=False, + ) + local_head = (head_res.stdout or "").strip() if head_res.returncode == 0 else None + + # Observe remote-tracking head without network when possible. + remote_head = None + for ref in ( + f"refs/remotes/origin/{branch_name}", + f"origin/{branch_name}", + f"refs/remotes/{remote}/{branch_name}", + f"{remote}/{branch_name}", + ): + rh = subprocess.run( + ["git", "-C", resolved_wt, "rev-parse", "--verify", "--quiet", ref], + capture_output=True, + text=True, + check=False, + ) + if rh.returncode == 0 and (rh.stdout or "").strip(): + remote_head = (rh.stdout or "").strip() + break + if remote_head is None: + # Fall back to caller's pin only for observation absence — assessment + # still requires pin==observed, so missing observation fails closed. + remote_head = None + + # Competing live locks (other issues / other worktrees). + competing_live = [] + for entry in issue_lock_store.list_live_locks(): + competing_live.append(entry) + + # Session pointers that claim this issue lock. + competing_sessions = [] + lock_dir = issue_lock_store.default_lock_dir() + lock_path = issue_lock_store.lock_file_path( + remote=remote, org=o, repo=r, issue_number=issue_number, lock_dir=lock_dir + ) + try: + for name in os.listdir(lock_dir): + if not name.startswith("session-") or not name.endswith(".json"): + continue + ptr = issue_lock_store.read_lock_file(os.path.join(lock_dir, name)) + if not ptr: + continue + ptr_lock = str(ptr.get("lock_file_path") or "").strip() + if not ptr_lock: + continue + try: + same = os.path.realpath(ptr_lock) == os.path.realpath(lock_path) + except OSError: + same = ptr_lock == lock_path + if not same: + continue + try: + sess_pid = int(str(name)[len("session-") : -len(".json")]) + except ValueError: + sess_pid = ptr.get("pid") + competing_sessions.append( + { + "pid": sess_pid, + "lock_file_path": ptr_lock, + "live": issue_lock_store.is_process_alive(sess_pid), + } + ) + except OSError: + pass + + # Best-effort workflow-lease scan: any live lock file whose work_lease is a + # non-author workflow lease on this issue/branch counts as active. + workflow_lease_active = False + for path in issue_lock_store.iter_lock_files(lock_dir): + rec = issue_lock_store.read_lock_file(path) + if not rec: + continue + lease = rec.get("work_lease") if isinstance(rec.get("work_lease"), dict) else {} + op = str(lease.get("operation_type") or "") + if op and op != issue_lock_store.AUTHOR_ISSUE_WORK_LEASE: + if rec.get("issue_number") == issue_number or str( + rec.get("branch_name") or "" + ) == branch_name: + if issue_lock_store.is_lease_live(rec): + workflow_lease_active = True + break + + repo_root = _canonical_local_git_root() + # permission_allowed reflects profile gate only — never ownership proof. + permission_allowed = True + + result = dirty_same_claimant_session_rebind.apply_dirty_same_claimant_session_rebind( + remote=remote, + org=o, + repo=r, + issue_number=issue_number, + branch_name=branch_name, + worktree_path=resolved_wt, + claimant_identity=identity, + claimant_profile=profile_name, + old_pid=old_pid, + expected_local_head=expected_local_head, + expected_remote_head=expected_remote_head, + expected_dirty_paths=list(expected_dirty_paths or []), + expected_fingerprints=dict(expected_fingerprints or {}), + existing_lock=existing, + current_identity=identity, + current_profile=profile_name, + role_kind=role_norm or role, + current_pid=os.getpid(), + current_branch=current_branch, + local_head=local_head, + remote_head=remote_head, + dirty_inventory=inv, + competing_live_locks=competing_live, + competing_sessions=competing_sessions, + workflow_lease_active=workflow_lease_active, + authorize_reconciler_execute=bool(authorize_reconciler_execute), + permission_allowed=permission_allowed, + repo_root=repo_root, + dry_run=bool(dry_run), + lock_dir=lock_dir, + ) + result["observed"] = { + "local_head": local_head, + "remote_head": remote_head, + "current_branch": current_branch, + "dirty_paths": inv.get("dirty_paths"), + "fingerprints": inv.get("fingerprints"), + "identity": identity, + "profile": profile_name, + "role_kind": role_norm, + } + return result + + return result + + @mcp.tool() def gitea_assess_work_issue_duplicate( issue_number: int, @@ -10839,6 +11409,9 @@ def _collect_branch_ownership_records( """ records: list[dict] = [] inventory_error = False + # #855 AC4: expired/stale reviewer-lease records eligible for an explicit + # reclaim decision, evaluated after the full ownership inventory is built. + reviewer_reclaim_candidates: list[tuple[dict, bool | None]] = [] target_branch = (branch or "").strip() if not target_branch: return {"records": records, "inventory_error": False} @@ -10987,15 +11560,28 @@ def _collect_branch_ownership_records( else: status = freshness_status reclaim_allowed = False - records.append( - _base_rec( - category=category, - status=status, - reclaim_allowed=reclaim_allowed, - role=role, - host=lease_host or host_n or host, - ) + rec = _base_rec( + category=category, + status=status, + reclaim_allowed=reclaim_allowed, + role=role, + host=lease_host or host_n or host, ) + records.append(rec) + # #855 AC4: a reviewer lease that is expired/stale (its owner + # gone) becomes a candidate for an explicit, fail-closed + # reclaim decision made once the full inventory is known. + if ( + role == "reviewer" + and status + in branch_cleanup_guard._RECLAIMABLE_REVIEWER_STATUSES + ): + owner_alive = ( + fr.get("owner_pid_alive") if isinstance(fr, dict) else None + ) + reviewer_reclaim_candidates.append( + (rec, owner_alive if isinstance(owner_alive, bool) else None) + ) except Exception: # O1: fail closed on control-plane inventory errors. inventory_error = True @@ -11066,6 +11652,44 @@ def _collect_branch_ownership_records( ) ) + # #855 AC4: decide, explicitly and fail-closed, whether any expired/stale + # reviewer lease may stop protecting an already-merged branch. This runs + # only after the full ownership inventory is built, so a competing active + # claimant (an active lease, author session, worktree binding, or active + # reviewer comment lease) is visible. An inventory failure keeps every + # reclaim candidate protective (reclaim_allowed stays False). + if reviewer_reclaim_candidates and not inventory_error: + pr_merged_state: bool | None = None + if pr_number is not None and auth and base_api: + try: + pr_live = api_request( + "GET", f"{base_api}/pulls/{int(pr_number)}", auth + ) + if isinstance(pr_live, dict) and pr_live: + pr_merged_state = bool( + pr_live.get("merged") or pr_live.get("merged_at") + ) + except Exception: + # Unknown merged state fails closed (candidate stays protective). + pr_merged_state = None + for cand_rec, owner_alive in reviewer_reclaim_candidates: + competing = any( + other is not cand_rec + and branch_cleanup_guard.is_active_ownership_status( + other.get("status") + ) + for other in records + ) + decision = branch_cleanup_guard.assess_expired_reviewer_lease_reclaim( + role=str(cand_rec.get("role")), + status=str(cand_rec.get("status")), + pr_merged=pr_merged_state, + owner_pid_alive=owner_alive, + competing_active_claimant=competing, + ) + cand_rec["reclaim_allowed"] = decision["reclaim_allowed"] + cand_rec["reclaim_decision"] = decision["decision"] + return {"records": records, "inventory_error": inventory_error} @@ -11128,6 +11752,7 @@ def gitea_reconcile_merged_cleanups( dry_run: bool = True, execute_confirmed: bool = False, limit: int = 50, + pr_number: int | None = None, remote: str = "dadeschools", host: str | None = None, org: str | None = None, @@ -11138,7 +11763,11 @@ def gitea_reconcile_merged_cleanups( Args: dry_run: Defaults to True. When True, only builds the reconciliation report. execute_confirmed: Must be True when dry_run=False. - limit: Max number of closed PRs to inspect. + limit: Max number of closed PRs to inspect (batch mode only; ignored when + ``pr_number`` is set). + pr_number: Optional exact merged PR selector (#855). When set, only that + PR is assessed/acted on (fail closed if missing, unmerged, or + ambiguous). When omitted, existing batch behaviour is preserved. remote: Known Gitea instance ('dadeschools' or 'prgs'). host: Override the Gitea host. org: Override the owner/organization. @@ -11173,11 +11802,120 @@ def gitea_reconcile_merged_cleanups( "audit_phase": audit_reconciliation_mode.current_phase(), } + # #855: optional exact PR pin. Fail closed before any inventory mutation. + exact_pr: int | None = None + if pr_number is not None: + try: + exact_pr = int(pr_number) + except (TypeError, ValueError): + return { + "success": False, + "performed": False, + "executed": False, + "dry_run": bool(dry_run), + "selection_mode": "exact_pr", + "selected_pr_number": pr_number, + "reasons": [ + f"pr_number={pr_number!r} is not a valid integer " + "(fail closed; no mutation)" + ], + "blocker_kind": "invalid_pr_number", + } + if exact_pr <= 0: + return { + "success": False, + "performed": False, + "executed": False, + "dry_run": bool(dry_run), + "selection_mode": "exact_pr", + "selected_pr_number": exact_pr, + "reasons": [ + f"pr_number={exact_pr} must be a positive integer " + "(fail closed; no mutation)" + ], + "blocker_kind": "invalid_pr_number", + } + h, o, r = _resolve(remote, host, org, repo) auth = _auth(h) base = repo_api_url(h, o, r) - closed_prs = api_get_all(f"{base}/pulls?state=closed", auth, limit=limit) - open_prs = api_get_all(f"{base}/pulls?state=open", auth) + + selection_mode = "batch" + closed_prs: list[dict] = [] + open_prs: list[dict] = [] + if exact_pr is not None: + selection_mode = "exact_pr" + try: + pr_live = api_request("GET", f"{base}/pulls/{exact_pr}", auth) + except Exception as exc: + return { + "success": False, + "performed": False, + "executed": False, + "dry_run": bool(dry_run), + "selection_mode": selection_mode, + "selected_pr_number": exact_pr, + "reasons": [ + f"PR #{exact_pr} could not be uniquely resolved " + f"(fail closed; no mutation): {_redact(str(exc))}" + ], + "blocker_kind": "pr_unresolvable", + } + if not isinstance(pr_live, dict) or not pr_live: + return { + "success": False, + "performed": False, + "executed": False, + "dry_run": bool(dry_run), + "selection_mode": selection_mode, + "selected_pr_number": exact_pr, + "reasons": [ + f"PR #{exact_pr} could not be uniquely resolved " + "(empty response; fail closed; no mutation)" + ], + "blocker_kind": "pr_unresolvable", + } + live_number = pr_live.get("number") + try: + live_number_int = int(live_number) if live_number is not None else None + except (TypeError, ValueError): + live_number_int = None + if live_number_int != exact_pr: + return { + "success": False, + "performed": False, + "executed": False, + "dry_run": bool(dry_run), + "selection_mode": selection_mode, + "selected_pr_number": exact_pr, + "reasons": [ + f"PR #{exact_pr} resolution is ambiguous or mismatched " + f"(live number={live_number!r}; fail closed; no mutation)" + ], + "blocker_kind": "pr_ambiguous", + } + if not (pr_live.get("merged") or pr_live.get("merged_at")): + return { + "success": False, + "performed": False, + "executed": False, + "dry_run": bool(dry_run), + "selection_mode": selection_mode, + "selected_pr_number": exact_pr, + "reasons": [ + f"PR #{exact_pr} is not merged " + "(exact-target cleanup requires a merged PR; " + "fail closed; no mutation)" + ], + "blocker_kind": "pr_not_merged", + } + closed_prs = [pr_live] + # Exact mode still needs open heads for remote-delete safety gates. + open_prs = api_get_all(f"{base}/pulls?state=open", auth) + else: + # Preserve historical call order (closed then open) for batch callers/tests. + closed_prs = api_get_all(f"{base}/pulls?state=closed", auth, limit=limit) + open_prs = api_get_all(f"{base}/pulls?state=open", auth) merged_closed: list[dict] = [] remote_branch_exists: dict[str, bool] = {} @@ -11204,6 +11942,13 @@ def gitea_reconcile_merged_cleanups( scratch_candidates = merged_cleanup_reconcile.discover_reviewer_scratch_worktrees( _canonical_local_git_root() ) + # #855: exact-target never inventories or mutates foreign PR scratch trees. + if exact_pr is not None: + scratch_candidates = [ + s + for s in scratch_candidates + if int(s.get("pr_number") or 0) == int(exact_pr) + ] active_reviewer_leases: dict[int, bool] = {} pr_states: dict[int, dict] = {} for scratch in scratch_candidates: @@ -11238,131 +11983,194 @@ def gitea_reconcile_merged_cleanups( active_reviewer_leases=active_reviewer_leases, pr_states=pr_states, ) + report["selection_mode"] = selection_mode + if exact_pr is not None: + report["selected_pr_number"] = exact_pr + # Fail closed if exact pin somehow produced other or zero entries. + entries = list(report.get("entries") or []) + entry_numbers = [] + for entry in entries: + try: + entry_numbers.append(int(entry.get("pr_number"))) + except (TypeError, ValueError): + entry_numbers.append(entry.get("pr_number")) + if entry_numbers != [exact_pr]: + return { + "success": False, + "performed": False, + "executed": False, + "dry_run": bool(dry_run), + "selection_mode": selection_mode, + "selected_pr_number": exact_pr, + "reasons": [ + f"exact PR #{exact_pr} selection produced unexpected " + f"candidate set {entry_numbers!r} " + "(fail closed; no mutation)" + ], + "blocker_kind": "exact_selection_mismatch", + "entries": entries, + } if dry_run: report["dry_run"] = True report["executed"] = False + # #851: surface planned lifecycle order so dry-run matches execute. + report["planned_execution_orders"] = { + str(entry.get("pr_number")): entry.get("planned_execution_order") or [] + for entry in (report.get("entries") or []) + } return {"success": True, "performed": False, **report} verify_preflight_purity( remote, task="reconcile_merged_cleanups", org=org, repo=repo ) actions: list[dict] = [] + project_root = _canonical_local_git_root() + + def _ownership_records_for_branch( + head_branch: str, pr_num_int: int | None + ) -> list[dict]: + ownership_bundle = _collect_branch_ownership_records( + remote=remote, + host=h, + org=o, + repo=r, + branch=head_branch, + pr_number=pr_num_int, + project_root=project_root, + auth=auth, + base_api=base, + ) + ownership_records = list(ownership_bundle.get("records") or []) + if ownership_bundle.get("inventory_error"): + ownership_records.append( + { + "category": ( + branch_cleanup_guard.OWNERSHIP_CATEGORY_INVENTORY_ERROR + ), + "status": "unknown", + "remote": remote, + "host": h, + "org": o, + "repo": r, + "branch": head_branch, + "reclaim_allowed": False, + "role": "inventory", + } + ) + return ownership_records + + def _attempt_owned_remote_delete( + *, + head_branch: str, + pr_num_int: int | None, + after_worktree_removal: bool = False, + ) -> dict: + """Fail-closed remote delete with live ownership reassessment (#851).""" + import urllib.parse + + ownership_records = _ownership_records_for_branch(head_branch, pr_num_int) + ownership = branch_cleanup_guard.assess_active_branch_ownership( + remote=remote, + org=o, + repo=r, + branch=head_branch, + host=h, + records=ownership_records, + ) + if ownership.get("block"): + return { + "action": "delete_remote_branch", + "branch": head_branch, + "success": False, + "performed": False, + "delete_acknowledged": False, + "verified_absent": False, + "blocker_kind": "active_branch_ownership", + "reasons": ownership.get("reasons") or [], + "blocking_categories": ownership.get("blocking_categories") or [], + "after_worktree_removal": after_worktree_removal, + "ownership_reassessed": after_worktree_removal, + } + + encoded = urllib.parse.quote(head_branch, safe="") + url = f"{base}/branches/{encoded}" + with _audited( + "delete_branch", + host=h, + remote=remote, + org=o, + repo=r, + target_branch=head_branch, + request_metadata={ + "branch": head_branch, + "source": "reconcile_merged_cleanups", + "ownership_checked": True, + "after_worktree_removal": after_worktree_removal, + }, + ): + api_request("DELETE", url, auth) + readback = _probe_remote_branch(h, o, r, auth, head_branch) + readback_assessment = branch_cleanup_guard.assess_post_delete_readback( + readback + ) + verified = bool(readback_assessment.get("verified_absent")) + return { + "action": "delete_remote_branch", + "branch": head_branch, + "success": bool(readback_assessment.get("ok")), + "performed": True, + "delete_acknowledged": True, + "verified_absent": verified, + "readback": readback_assessment.get("readback"), + "reasons": readback_assessment.get("reasons") or [], + "after_worktree_removal": after_worktree_removal, + "ownership_reassessed": after_worktree_removal, + } + for entry in report.get("entries") or []: head_branch = entry.get("head_branch") or "" remote_assessment = entry.get("remote_branch") or {} local_assessment = entry.get("local_worktree") or {} + pr_num = entry.get("pr_number") + try: + pr_num_int = int(pr_num) if pr_num is not None else None + except (TypeError, ValueError): + pr_num_int = None - if remote_assessment.get("safe_to_delete_remote"): - import urllib.parse - - pr_num = entry.get("pr_number") - try: - pr_num_int = int(pr_num) if pr_num is not None else None - except (TypeError, ValueError): - pr_num_int = None - ownership_bundle = _collect_branch_ownership_records( - remote=remote, - host=h, - org=o, - repo=r, - branch=head_branch, - pr_number=pr_num_int, - project_root=_canonical_local_git_root(), - auth=auth, - base_api=base, - ) - ownership_records = list(ownership_bundle.get("records") or []) - if ownership_bundle.get("inventory_error"): - ownership_records.append( - { - "category": ( - branch_cleanup_guard.OWNERSHIP_CATEGORY_INVENTORY_ERROR - ), - "status": "unknown", - "remote": remote, - "host": h, - "org": o, - "repo": r, - "branch": head_branch, - "reclaim_allowed": False, - "role": "inventory", - } - ) - ownership = branch_cleanup_guard.assess_active_branch_ownership( - remote=remote, - org=o, - repo=r, - branch=head_branch, - host=h, - records=ownership_records, - ) - if ownership.get("block"): - actions.append( - { - "action": "delete_remote_branch", - "branch": head_branch, - "success": False, - "performed": False, - "delete_acknowledged": False, - "verified_absent": False, - "blocker_kind": "active_branch_ownership", - "reasons": ownership.get("reasons") or [], - "blocking_categories": ownership.get( - "blocking_categories" - ) - or [], - } - ) - continue - - encoded = urllib.parse.quote(head_branch, safe="") - url = f"{base}/branches/{encoded}" - with _audited( - "delete_branch", - host=h, - remote=remote, - org=o, - repo=r, - target_branch=head_branch, - request_metadata={ - "branch": head_branch, - "source": "reconcile_merged_cleanups", - "ownership_checked": True, - }, - ): - api_request("DELETE", url, auth) - readback = _probe_remote_branch(h, o, r, auth, head_branch) - readback_assessment = branch_cleanup_guard.assess_post_delete_readback( - readback - ) - verified = bool(readback_assessment.get("verified_absent")) - actions.append( - { - "action": "delete_remote_branch", - "branch": head_branch, - "success": bool(readback_assessment.get("ok")), - "performed": True, - "delete_acknowledged": True, - "verified_absent": verified, - "readback": readback_assessment.get("readback"), - "reasons": readback_assessment.get("reasons") or [], - } - ) - + # #851 lifecycle: when the target worktree is independently safe, remove + # it first so worktree_binding ownership does not permanently strand + # both the worktree and the remote branch. Never skip worktree removal + # merely because remote delete would be blocked by that binding. + # Ownership protection for remote delete remains fail-closed below. + worktree_removed = False if local_assessment.get("safe_to_remove_worktree"): result = merged_cleanup_reconcile.remove_local_worktree( - _canonical_local_git_root(), + project_root, head_branch, worktree_path=local_assessment.get("worktree_path"), ) actions.append({"action": "remove_local_worktree", **result}) + # Idempotent resume: absent worktree is already gone. + msg = (result.get("message") or "").lower() + worktree_removed = bool(result.get("success")) or ( + "not found" in msg + ) + + if remote_assessment.get("safe_to_delete_remote"): + actions.append( + _attempt_owned_remote_delete( + head_branch=head_branch, + pr_num_int=pr_num_int, + after_worktree_removal=worktree_removed, + ) + ) for scratch in report.get("reviewer_scratch_entries") or []: if not scratch.get("safe_to_remove_worktree"): continue result = merged_cleanup_reconcile.remove_reviewer_scratch_worktree( - _canonical_local_git_root(), scratch.get("worktree_path") or "" + project_root, scratch.get("worktree_path") or "" ) actions.append({"action": "remove_reviewer_scratch_worktree", **result}) @@ -11598,6 +12406,7 @@ def gitea_audit_worktree_cleanup( org: str | None = None, repo: str | None = None, ttl_hours: float = worktree_cleanup_audit.DEFAULT_TTL_HOURS, + merged_pr_limit: int = 200, ) -> dict: """Read-only: classify every session-owned worktree under ``branches/`` (#401). @@ -11608,17 +12417,26 @@ def gitea_audit_worktree_cleanup( the active issue-lock branch is read from the local lock file and treated as active work. Deletes nothing and mutates no Gitea state. - Fails closed if the live open-PR list cannot be fetched: without it, - removability cannot be proven, so no candidates are returned. + Merged PRs are fetched as well, so an issue worktree can be linked to the + PR that owns its branch (#858). Such a worktree only becomes removable + when that owning PR is unambiguous and merged, the worktree head is + already contained in authoritative master, and nothing else protects it — + no open or competing PR, lease, issue lock, live session, dirty file, or + protected/control checkout. Anything unproven keeps it classified as + active issue work. + + Fails closed if the live open-PR list, the merged-PR list, or the + control-plane lease state cannot be read: without them removability + cannot be proven, so no candidates are returned. Args: remote: Known instance — 'dadeschools' or 'prgs'. host: Override the Gitea host. org: Override the owner/organization. repo: Override the repository name. - ttl_hours: Age (hours) after which a clean issue/conflict-fix - worktree becomes stale-removable (default from - GITEA_WORKTREE_TTL_HOURS). + ttl_hours: Age (hours) after which a clean conflict-fix worktree + becomes stale-removable (default from GITEA_WORKTREE_TTL_HOURS). + merged_pr_limit: Max closed PRs scanned for merged-PR ownership. Returns: dict with per-worktree classifications, counts, removable @@ -11654,22 +12472,84 @@ def gitea_audit_worktree_cleanup( if (pr.get("head") or {}).get("ref") } + # #858: merged PRs are the ownership evidence that lets a landed issue + # worktree stop being reported as active work. Without them the audit can + # never agree with the PR-scoped reconciler, so treat a fetch failure the + # same way an open-PR fetch failure is treated: fail closed. + try: + closed_prs = api_get_all( + f"{repo_api_url(h, o, r)}/pulls?state=closed", auth, limit=merged_pr_limit + ) + except Exception as exc: + return { + "success": False, + "performed": False, + "open_pr_state_verified": True, + "merged_pr_state_verified": False, + "reasons": [ + "could not fetch merged PRs; worktree ownership unverified " + f"(fail closed): {_redact(str(exc))}" + ], + } + merged_prs = [pr for pr in closed_prs if (pr.get("merged") or pr.get("merged_at"))] + pr_index = worktree_cleanup_audit.build_pr_index(list(open_prs) + merged_prs) + + # #858: the auditor already accepted lease evidence but nothing ever + # supplied it, so every worktree looked unleased. Removability is now + # reachable for issue worktrees, so authoritative control-plane leases + # must be readable or the audit fails closed. + db, lease_errs = _control_plane_db_or_error() + if db is None: + return { + "success": False, + "performed": False, + "open_pr_state_verified": True, + "merged_pr_state_verified": True, + "lease_state_verified": False, + "reasons": [ + "could not read control-plane leases; worktree protection " + "unverified (fail closed)", + *lease_errs, + ], + } + lease_result = lease_lifecycle.list_active_leases( + db, remote=remote, org=o, repo=r, include_non_active=False, limit=500 + ) + leased_issue_numbers: set[int] = set() + live_session_paths: set[str] = set() + for lease in lease_result.get("leases") or []: + if lease.get("work_kind") == "issue" and lease.get("work_number") is not None: + try: + leased_issue_numbers.add(int(lease["work_number"])) + except (TypeError, ValueError): + pass + if lease.get("worktree_path"): + live_session_paths.add(str(lease["worktree_path"])) + active_issue_branches: set[str] = set() lock = merged_cleanup_reconcile.read_issue_lock(ISSUE_LOCK_FILE) if lock and lock.get("branch_name"): active_issue_branches.add(str(lock["branch_name"]).strip()) + master_ref = f"{remote}/master" if remote in REMOTES else "origin/master" report = worktree_cleanup_audit.audit_branches_directory( _canonical_local_git_root(), open_pr_branches=open_pr_branches, active_issue_branches=active_issue_branches, now=datetime.now(timezone.utc), ttl_hours=ttl_hours, + pr_index=pr_index, + leased_issue_numbers=leased_issue_numbers, + live_session_paths=live_session_paths, + master_ref=master_ref, ) return { "success": True, "performed": False, "open_pr_state_verified": True, + "merged_pr_state_verified": True, + "lease_state_verified": True, + "master_ref": master_ref, "task_mode": "work-issue", **report, } @@ -18504,10 +19384,59 @@ def gitea_update_pr_branch_by_merge( prepared_verdict_head_sha=live_pr_head, ) - return { - "success": True, + # #871: the remote head is now advanced; the durable linked-issue lock must + # be advanced with it, or a later dead-session recovery can never prove + # ownership at the new head. This runs AFTER the successful remote update, so + # a failure here is a *partial* lifecycle failure — the remote moved but the + # durable state did not — and must never be reported as a full success. + claimant = _work_lease_claimant(h) + matched_issue = ownership.get("matched_issue") + synced_at = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ") + lock_refresh: dict = { + "refreshed": False, + "reasons": ["durable lock head refresh was not attempted"], + } + if new_head and matched_issue and source_branch and (wt or None): + try: + lock_refresh = issue_lock_store.apply_durable_lock_head_refresh( + remote=remote, + org=o, + repo=r, + issue_number=int(matched_issue), + branch_name=source_branch, + worktree_path=wt, + pr_number=pr_number, + identity=claimant.get("username"), + profile=claimant.get("profile"), + current_pid=os.getpid(), + expected_old_head=live_pr_head, + new_head=new_head, + synced_at=synced_at, + base_head=live_base_head, + ) + except Exception as exc: + lock_refresh = { + "refreshed": False, + "reasons": [ + f"durable lock head refresh raised (fail closed): {_redact(str(exc))}" + ], + } + else: + lock_refresh = { + "refreshed": False, + "reasons": [ + "durable lock head refresh could not run: missing new head, " + "linked issue, source branch, or worktree binding" + ], + } + + durable_refreshed = bool(lock_refresh.get("refreshed")) + base_result = { "performed": True, "mutation_allowed": True, + "durable_lock_refreshed": durable_refreshed, + "durable_lock_refresh": lock_refresh, + "fully_synchronized": durable_refreshed, "style": "merge", "force_push": False, "rebase": False, @@ -18529,19 +19458,37 @@ def gitea_update_pr_branch_by_merge( "prepared_verdict_invalidated": transition.get( "prepared_verdict_invalidated" ), - "recommended_next_action": transition.get( - "recommended_next_action", - pr_sync_status.ACTION_FRESH_REVIEW_REQUIRED, - ), "transition": transition, "role_kind": role, "profile_name": profile.get("profile_name"), "worktree_path": wt or None, - "reasons": list(transition.get("reasons") or []) + [ - "update-by-merge completed via native Gitea API (style=merge only)" - ], } + if durable_refreshed: + base_result["success"] = True + base_result["recommended_next_action"] = transition.get( + "recommended_next_action", + pr_sync_status.ACTION_FRESH_REVIEW_REQUIRED, + ) + base_result["reasons"] = list(transition.get("reasons") or []) + [ + "update-by-merge completed via native Gitea API (style=merge only)", + f"durable linked-issue lock #{matched_issue} head refreshed to " + f"{new_head} (verified by read-after-write)", + ] + return base_result + + # Partial lifecycle failure: the remote advanced but the durable lock did + # not. Do NOT report a fully successful synchronization (#871). + base_result["success"] = False + base_result["partial_lifecycle_failure"] = True + base_result["recommended_next_action"] = pr_sync_status.ACTION_BLOCKED + base_result["reasons"] = list(lock_refresh.get("reasons") or []) + [ + f"PARTIAL LIFECYCLE FAILURE: PR #{pr_number} remote head advanced to " + f"{new_head} but the durable linked-issue lock head was not refreshed; " + "the synchronization is NOT complete", + ] + return base_result + @mcp.tool() def gitea_assess_conflict_fix_push( @@ -21104,6 +22051,156 @@ def gitea_workflow_dashboard( return payload +@mcp.tool() +def gitea_request_mcp_restart( + remote: str = "dadeschools", + host: str | None = None, + org: str | None = None, + repo: str | None = None, + dry_run: bool = True, + request_override: bool = False, + session_id: str | None = None, + limit: int = 200, +) -> dict: + """Evaluate a proposed MCP restart and return an impact preview (#658). + + Central restart coordinator: gathers live control-plane state (sessions, + leases/locks, in-flight issue/PR work, mutations, worktrees) and returns a + blast-radius impact report with a ``safe`` / ``unsafe`` / ``override`` + verdict, so the console (#642/#652) and operators can see what a restart + would disrupt *before* any concurrent LLM work is destroyed. + + This tool is **dry-run and never restarts anything.** The mutative apply + path is a separate child gated by a drain proof (non-goal here); calling + with ``dry_run=False`` still performs no restart and reports that apply is + not yet available. + + Operator override authority is read from the process environment + (``GITEA_OPERATOR_RESTART_OVERRIDE_AUTHORIZATION``), never self-asserted by + the requesting session: ``request_override`` only expresses caller intent + and takes effect solely when that environment authorization is present. + + Fails closed: if the control-plane inventory cannot be completed, the + verdict is ``unsafe`` / deny (an incomplete evaluation must never green-light + a restart). + """ + read_block = _profile_operation_gate("gitea.read") + if read_block: + return { + "success": False, + "read_only": True, + "dry_run": True, + "restart_performed": False, + "reasons": read_block, + "permission_report": _permission_block_report("gitea.read"), + } + + try: + h, o, r = _resolve(remote, host, org, repo) + except ValueError as exc: + return { + "success": False, + "read_only": True, + "dry_run": True, + "restart_performed": False, + "reasons": [str(exc)], + } + + inventory_complete = True + incomplete_reasons: list[str] = [] + sessions: list[dict] = [] + leases: list[dict] = [] + terminal_lock: dict | None = None + + db, db_errs = _control_plane_db_or_error() + if db is None: + inventory_complete = False + incomplete_reasons.extend( + db_errs or ["control-plane DB unavailable; cannot evaluate restart"] + ) + else: + try: + sessions = db.list_sessions(statuses=("active",), limit=max(1, int(limit))) + except Exception as exc: # noqa: BLE001 + inventory_complete = False + incomplete_reasons.append( + f"session inventory failed: {_redact(str(exc))}" + ) + try: + lease_result = lease_lifecycle.list_active_leases( + db, + remote=remote if remote in REMOTES else remote, + org=o, + repo=r, + role=None, + include_non_active=False, + limit=max(1, int(limit)), + ) + leases = list(lease_result.get("leases") or []) + except Exception as exc: # noqa: BLE001 + inventory_complete = False + incomplete_reasons.append( + f"lease inventory failed: {_redact(str(exc))}" + ) + try: + terminal = db.get_active_terminal_lock( + remote=remote if remote in REMOTES else remote, + org=o, + repo=r, + ) + if terminal: + terminal_lock = dict(terminal) + except Exception as exc: # noqa: BLE001 + inventory_complete = False + incomplete_reasons.append( + f"terminal lock lookup failed: {_redact(str(exc))}" + ) + + profile = get_profile() + profile_name = (profile.get("profile_name") or "").strip() or "session" + sid = (session_id or "").strip() or f"{profile_name}-{os.getpid()}" + + # Override authority is read from the environment only — a worker session + # cannot set an env var for an already-running daemon, so it cannot be + # self-asserted the way a tool argument could (#630/#710 F1 pattern). + operator_authorized = bool( + (os.environ.get("GITEA_OPERATOR_RESTART_OVERRIDE_AUTHORIZATION") or "").strip() + ) + operator_override = bool(request_override and operator_authorized) + + inventory = { + "sessions": sessions, + "leases": leases, + "terminal_lock": terminal_lock, + "inventory_complete": inventory_complete, + "incomplete_reasons": incomplete_reasons, + } + + report = restart_coordinator.evaluate_restart_impact( + inventory, + operator_override=operator_override, + requesting_session_id=sid, + dry_run=True, # coordinator is always analysis-only (#658) + ) + + payload = report.as_dict() + payload["success"] = True + payload["read_only"] = True + payload["remote"] = remote + payload["org"] = o + payload["repo"] = r + payload["requesting_session_id"] = sid + payload["operator_override_requested"] = bool(request_override) + payload["operator_override_authorized"] = operator_authorized + payload["apply_supported"] = False + if not dry_run: + payload["reasons"] = list(payload.get("reasons") or []) + [ + "apply requested but not supported: sanctioned restart apply is " + "gated by a drain proof (separate child); no restart performed (#658)" + ] + return payload + + @mcp.tool() def gitea_inspect_workflow_lease( lease_id: str, diff --git a/issue_lock_provenance.py b/issue_lock_provenance.py index 87ee38f..b79d8ad 100644 --- a/issue_lock_provenance.py +++ b/issue_lock_provenance.py @@ -16,11 +16,18 @@ ISSUE_LOCK_FILE = os.environ.get("GITEA_ISSUE_LOCK_FILE", "/tmp/gitea_issue_lock SOURCE_LOCK_ISSUE = "gitea_lock_issue" SOURCE_LOCK_ADOPTION = "gitea_lock_issue_adoption" SOURCE_OPERATOR_OVERRIDE = "operator_override" +SOURCE_RECOVER_DIRTY_ORPHANED = "gitea_recover_dirty_orphaned_issue_worktree" +# #864: dirty-preserving same-claimant author-session rebind (dead owner PID). +SOURCE_DIRTY_SAME_CLAIMANT_REBIND = ( + "gitea_rebind_dirty_same_claimant_author_session" +) SANCTIONED_LOCK_SOURCES = frozenset({ SOURCE_LOCK_ISSUE, SOURCE_LOCK_ADOPTION, SOURCE_OPERATOR_OVERRIDE, + SOURCE_RECOVER_DIRTY_ORPHANED, + SOURCE_DIRTY_SAME_CLAIMANT_REBIND, }) _OPERATOR_OVERRIDE_ENV = "GITEA_ISSUE_LOCK_OPERATOR_OVERRIDE" diff --git a/issue_lock_recovery.py b/issue_lock_recovery.py index ecc99c6..ae0cc35 100644 --- a/issue_lock_recovery.py +++ b/issue_lock_recovery.py @@ -85,6 +85,12 @@ HEAD_RELATION_STRICT_DESCENDANT = "strict_descendant" # #772: an unpublished claim has no recorded head to compare against at all, so # its head is measured against the base the branch was cut from instead. HEAD_RELATION_DESCENDS_FROM_BASE = "descends_from_recorded_base" +# #871: the remote/PR head advanced *past* the recorded head via a sanctioned +# merge-based branch synchronization (``gitea_update_pr_branch_by_merge``) while +# the local worktree stayed at the recorded head. This is the inverse of the +# #768 descendant relation — here the *remote* strictly descends the local head, +# and only because a base was merged into the branch, proven server-side. +HEAD_RELATION_REMOTE_MERGE_SYNCED = "remote_merge_synced" # Which body of evidence a recovery was decided on (#772 AC10). These are not # interchangeable: a published claim proves ownership against a remote/PR head, @@ -266,6 +272,70 @@ def _assess_base_descendancy( ] +def _assess_remote_merge_synced( + sync_provenance: Mapping[str, Any] | None, + *, + recorded_head: str, + remote_head: str, +) -> tuple[bool, list[str]]: + """Did ``remote_head`` advance past ``recorded_head`` via a sanctioned + merge-based branch sync (#871)? + + ``sync_provenance`` is the server-side git observation from + ``issue_lock_worktree.read_merge_sync_provenance``. Its own + ``prior_head_sha`` / ``synced_head_sha`` are re-checked against the heads + this assessment is actually reasoning about, so an observation taken for some + other pair of commits — stale, mismatched, or hand-built — can never + authorize recovery. This is the inverse of ``_assess_strict_descendant``: the + recorded head is the ancestor and the *remote* head is the descendant, and it + is accepted only because the remote head is a base-into-branch merge that + preserved the branch mainline back to the recorded head. + + Returns ``(proven, notes)``. Notes name the exact missing element so a + refused caller sees why, never a bare "unproven". + """ + if not isinstance(sync_provenance, Mapping): + return False, [ + "no server-derived merge-sync provenance observation was available; a " + "remote head ahead of the recorded head cannot be accepted" + ] + + probe_prior = _text(sync_provenance.get("prior_head_sha")) + probe_synced = _text(sync_provenance.get("synced_head_sha")) + if probe_prior != recorded_head or probe_synced != remote_head: + return False, [ + f"merge-sync observation covers {probe_prior or 'unknown'} -> " + f"{probe_synced or 'unknown'}, not the heads under assessment " + f"({recorded_head} -> {remote_head})" + ] + if not sync_provenance.get("probe_ok"): + return False, ( + list(sync_provenance.get("reasons") or []) + or ["merge-sync provenance probe did not complete; provenance unproven"] + ) + if not sync_provenance.get("prior_is_ancestor"): + return False, [ + f"recorded head {recorded_head} is not an ancestor of remote head " + f"{remote_head}; a rewritten or force-moved head cannot be recovered" + ] + if not sync_provenance.get("is_merge_sync"): + return False, ( + list(sync_provenance.get("reasons") or []) + or [ + f"remote head {remote_head} is not a sanctioned merge-based sync " + f"of the base into the branch above {recorded_head}" + ] + ) + + proof = _text(sync_provenance.get("proof")) or ( + f"{remote_head} merged the base into the branch above {recorded_head}" + ) + return True, [ + f"remote head {remote_head} advanced past recorded head {recorded_head} " + f"via a sanctioned merge-based branch sync ({proof})" + ] + + def assess_dead_session_lock_recovery( existing_lock: Mapping[str, Any] | None, *, @@ -290,6 +360,7 @@ def assess_dead_session_lock_recovery( remote_branch_exists: bool | None = None, recorded_base_sha: str | None = None, base_ancestry: Mapping[str, Any] | None = None, + sync_provenance: Mapping[str, Any] | None = None, ) -> dict[str, Any]: """Decide whether a dead-session author lock may be natively recovered. @@ -469,19 +540,39 @@ def assess_dead_session_lock_recovery( head_relation = HEAD_RELATION_STRICT_DESCENDANT ancestry_proof = notes[0] if notes else None else: - reasons.append( - f"local head {local_head} does not match remote branch head " - f"{remote_head}" + # #871: the reverse relation — the remote head advanced past + # the recorded/local head via a sanctioned merge-based branch + # sync while the local worktree stayed put. Accepted only on + # server-proven merge-sync provenance, never a caller claim. + synced, sync_notes = _assess_remote_merge_synced( + sync_provenance, + recorded_head=local_head, + remote_head=remote_head, ) - reasons.extend(notes) + if synced: + head_relation = HEAD_RELATION_REMOTE_MERGE_SYNCED + ancestry_proof = sync_notes[0] if sync_notes else None + else: + reasons.append( + f"local head {local_head} does not match remote branch " + f"head {remote_head}" + ) + reasons.extend(notes) + reasons.extend(sync_notes) evidence["recorded_base"] = recorded_base or None evidence["local_head"] = local_head or None evidence["remote_head"] = remote_head or None # ``recorded_head`` is the head recovery is being measured against; # ``accepted_head`` is the head this recovery actually adopts. They differ # only in the descendant case, and downstream gates need both (#768 AC2/AC7). + # #871: in the merge-sync case the branch/PR already carries the synced + # remote head, so that is the head recovery adopts; the local worktree stays + # at the ancestor recorded head. evidence["recorded_head"] = remote_head or None - evidence["accepted_head"] = local_head or None + if head_relation == HEAD_RELATION_REMOTE_MERGE_SYNCED: + evidence["accepted_head"] = remote_head or None + else: + evidence["accepted_head"] = local_head or None evidence["head_relation"] = head_relation evidence["ancestry_proof"] = ancestry_proof @@ -493,12 +584,17 @@ def assess_dead_session_lock_recovery( # contradictory; re-stating it as a head mismatch would only obscure why. if not unpublished and local_head and pr_head != local_head: # A descendant recovery has not been published yet, so the open PR - # legitimately still points at the recorded head. Any other - # disagreement is a real mismatch. + # legitimately still points at the recorded head. A merge-sync + # recovery's PR legitimately sits at the advanced remote head. Any + # other disagreement is a real mismatch. if not ( head_relation == HEAD_RELATION_STRICT_DESCENDANT and remote_head and pr_head == remote_head + ) and not ( + head_relation == HEAD_RELATION_REMOTE_MERGE_SYNCED + and remote_head + and pr_head == remote_head ): reasons.append( f"open PR #{pr_number} head {pr_head} does not match local head " @@ -619,7 +715,11 @@ def assess_dead_session_lock_recovery( ) if ( head_relation - in (HEAD_RELATION_STRICT_DESCENDANT, HEAD_RELATION_DESCENDS_FROM_BASE) + in ( + HEAD_RELATION_STRICT_DESCENDANT, + HEAD_RELATION_DESCENDS_FROM_BASE, + HEAD_RELATION_REMOTE_MERGE_SYNCED, + ) and ancestry_proof ): proof.append(ancestry_proof) @@ -697,6 +797,16 @@ def owning_pr_recovery_evidence( return None if accepted_head != local_head: return None + elif relation == HEAD_RELATION_REMOTE_MERGE_SYNCED: + # #871: the PR already sits at the advanced remote head; the local + # worktree is the ancestor the merge preserved. The head the open PR + # shows and the head recovery adopts are both the synced remote head. + if not remote_head or pr_head != remote_head: + return None + if accepted_head != remote_head: + return None + if not local_head or local_head == remote_head: + return None else: return None try: @@ -765,6 +875,18 @@ def recovered_owning_pr_from_lock( return None if not accepted_head or accepted_head == recorded_head: return None + elif relation == HEAD_RELATION_REMOTE_MERGE_SYNCED: + # #871: PR sits at the advanced remote head, which is both the recorded + # measured-against head and the adopted head; the local worktree is the + # ancestor the merge preserved. + remote_head = _text(record.get("remote_head")) + local_head = _text(record.get("local_head")) + if not remote_head or pr_head != remote_head: + return None + if accepted_head and accepted_head != remote_head: + return None + if not local_head or local_head == remote_head: + return None else: return None try: diff --git a/issue_lock_store.py b/issue_lock_store.py index 713fa2a..d011c1c 100644 --- a/issue_lock_store.py +++ b/issue_lock_store.py @@ -169,6 +169,7 @@ def bind_session_lock( *, expected_generation: int | None = None, renewal_sanctioned: bool = False, + recovery_sanctioned: bool = False, ) -> str: """Persist a keyed lock and bind it to the current process session. @@ -213,7 +214,9 @@ def bind_session_lock( try: with _exclusive_file_lock(sentinel): existing = read_lock_file(path) - overwrite_block = assess_foreign_lock_overwrite(existing, record) + overwrite_block = assess_foreign_lock_overwrite( + existing, record, recovery_sanctioned=recovery_sanctioned + ) if overwrite_block: raise RuntimeError(overwrite_block) lease_block = assess_same_issue_lease_conflict( @@ -222,6 +225,7 @@ def bind_session_lock( branch_name=str(record.get("branch_name") or ""), worktree_path=str(record.get("worktree_path") or ""), renewal_sanctioned=renewal_sanctioned, + recovery_sanctioned=recovery_sanctioned, ) if lease_block: raise RuntimeError(lease_block) @@ -380,7 +384,18 @@ def assess_lock_freshness( pid = lock_data.get("session_pid") if pid is None: pid = lock_data.get("pid") - pid_alive = is_process_alive(pid) if pid is not None else False + if pid is None: + pid = lock_data.get("owner_pid") + pid_missing = pid is None or str(pid).strip() == "" + try: + pid_int = int(pid) if not pid_missing else None + if pid_int is not None and pid_int <= 0: + pid_missing = True + pid_int = None + except (TypeError, ValueError): + pid_missing = True + pid_int = None + pid_alive = is_process_alive(pid_int) if pid_int is not None else False if expires_at and expires_at <= current: return { @@ -389,15 +404,36 @@ def assess_lock_freshness( "stale": True, "reason": f"lease expired at {expires_at.isoformat()}", "pid_alive": pid_alive, + "pid_missing": pid_missing, } - if pid is not None and not pid_alive: + # #860: a PID-less lock must never be considered live merely because + # expiration / heartbeat fields are absent. Missing PID is insufficient + # evidence of a live owner; treat as malformed/stale so recovery routes + # can evaluate corroborating pins instead of blocking on a false live flag. + if pid_missing: + return { + "status": "malformed", + "live": False, + "stale": True, + "reason": ( + "lock has no usable session pid; cannot prove live ownership " + "(PID-less locks are never live by missing expiry alone)" + ), + "pid_alive": False, + "pid_missing": True, + "heartbeat_at": heartbeat_at.isoformat() if heartbeat_at else None, + "expires_at": expires_at.isoformat() if expires_at else None, + } + + if pid_int is not None and not pid_alive: return { "status": "stale", "live": False, "stale": True, - "reason": f"owner pid {pid} is not alive", + "reason": f"owner pid {pid_int} is not alive", "pid_alive": False, + "pid_missing": False, } return { @@ -406,6 +442,7 @@ def assess_lock_freshness( "stale": False, "reason": "lock heartbeat and lease are fresh", "pid_alive": pid_alive, + "pid_missing": False, "heartbeat_at": heartbeat_at.isoformat() if heartbeat_at else None, "expires_at": expires_at.isoformat() if expires_at else None, } @@ -486,6 +523,7 @@ def assess_same_issue_lease_conflict( worktree_path: str, operation_type: str = AUTHOR_ISSUE_WORK_LEASE, renewal_sanctioned: bool = False, + recovery_sanctioned: bool = False, now: datetime | None = None, ) -> str | None: """Return a fail-closed error when a competing live lease blocks acquisition. @@ -517,6 +555,8 @@ def assess_same_issue_lease_conflict( existing_branch == branch_name and _same_realpath(str(existing_worktree or ""), worktree_path) ) + if recovery_sanctioned and existing_issue == issue_number and existing_branch == branch_name: + return None if is_lease_expired(existing_lock, now=now): # #760 AC1/AC2: exact-owner renewal is a different disposition from # foreign takeover and is evaluated first. Before this, both branches @@ -547,10 +587,26 @@ def assess_same_issue_lease_conflict( ) +def _lock_claimant(lock: dict[str, Any] | None) -> dict[str, str]: + if not isinstance(lock, dict): + return {} + claimant = lock.get("claimant") + if not isinstance(claimant, dict): + lease = lock.get("work_lease") + claimant = lease.get("claimant") if isinstance(lease, dict) else None + if not isinstance(claimant, dict): + return {} + return { + "username": str(claimant.get("username") or ""), + "profile": str(claimant.get("profile") or ""), + } + + def assess_foreign_lock_overwrite( existing_lock: dict[str, Any] | None, incoming_lock: dict[str, Any], *, + recovery_sanctioned: bool = False, now: datetime | None = None, ) -> str | None: """Block writes that would clobber an unrelated live lease on the same key.""" @@ -565,8 +621,31 @@ def assess_foreign_lock_overwrite( ) if same_issue and same_branch and same_worktree: return None - if not is_lease_live(existing_lock, now=now): + + existing_claimant = _lock_claimant(existing_lock) + incoming_claimant = _lock_claimant(incoming_lock) + same_claimant = ( + bool(existing_claimant.get("username")) + and existing_claimant.get("username") == incoming_claimant.get("username") + and existing_claimant.get("profile") == incoming_claimant.get("profile") + ) + + if recovery_sanctioned and same_issue and same_branch and same_claimant: return None + + if not is_lease_live(existing_lock, now=now): + # #860 F8: A non-live or PID-less lock still blocks foreign overwrite + # unless same claimant or sanctioned reclaim is proven. + if not same_claimant and same_issue: + reclaim = assess_expired_lock_reclaim(existing_lock, now=now) + if not reclaim.get("reclaim_allowed"): + return ( + "Refusing foreign overwrite of non-live issue lock " + f"(issue #{existing_lock.get('issue_number')}, owner '{existing_claimant.get('username')}') " + "without sanctioned reclaim proof (fail closed)" + ) + return None + return ( "Refusing to overwrite a live foreign issue lock " f"(issue #{existing_lock.get('issue_number')}, " @@ -737,4 +816,308 @@ def format_lock_proof( parts.append("lock released") elif released is False: parts.append("lock retained") - return "; ".join(parts) \ No newline at end of file + return "; ".join(parts) + + +# ── #871: durable linked-issue lock head refresh after branch synchronization ── +_FULL_SHA_RE = re.compile(r"^[0-9a-f]{40}$", re.IGNORECASE) + +# Provenance recorded on the lock when the head is refreshed by a sanctioned +# merge-based branch synchronization (``gitea_update_pr_branch_by_merge``). +LOCK_HEAD_REFRESH_PROVENANCE_MERGE_SYNC = "gitea_update_pr_branch_by_merge" + + +def _norm_sha(value: Any) -> str | None: + text = str(value or "").strip().lower() + return text if _FULL_SHA_RE.match(text) else None + + +def _lock_claimant_view(lock: dict[str, Any] | None) -> dict[str, Any]: + if not isinstance(lock, dict): + return {} + claimant = lock.get("claimant") + if not isinstance(claimant, dict): + lease = lock.get("work_lease") + claimant = lease.get("claimant") if isinstance(lease, dict) else None + return dict(claimant) if isinstance(claimant, dict) else {} + + +def assess_durable_lock_head_refresh( + existing_lock: dict[str, Any] | None, + *, + remote: str, + org: str, + repo: str, + issue_number: int, + branch_name: str, + worktree_path: str, + pr_number: int | None, + identity: str | None, + profile: str | None, + current_pid: int | None, + expected_old_head: str | None, + new_head: str | None, + base_head: str | None = None, +) -> dict[str, Any]: + """Fail-closed assessment for refreshing a durable lock's recorded head (#871). + + A successful ``gitea_update_pr_branch_by_merge`` advances the *remote* PR head + but must also advance the durable linked-issue lock so a later dead-session + recovery can prove ownership. This decides whether that refresh is permitted; + it mutates nothing. + + Every element of durable ownership is re-verified against the persisted lock — + repository, issue, branch, worktree, claimant identity/profile, and the live + owning session — and the recorded head is compare-and-swapped: the lock's + currently recorded synced head (if any) must equal ``expected_old_head``, so a + lock whose head or provenance changed concurrently is never overwritten. + """ + reasons: list[str] = [] + old = _norm_sha(expected_old_head) + new = _norm_sha(new_head) + evidence: dict[str, Any] = { + "issue_number": issue_number, + "branch_name": branch_name, + "worktree_path": worktree_path, + "pr_number": pr_number, + "expected_old_head": old, + "new_head": new, + "base_head": _norm_sha(base_head), + } + + if not isinstance(existing_lock, dict) or not existing_lock: + reasons.append("no durable lock exists for this issue; nothing to refresh") + return {"allowed": False, "reasons": reasons, "evidence": evidence, + "expected_generation": 0} + + lock = dict(existing_lock) + evidence["current_generation"] = lock_generation(lock) + + if lock.get("issue_number") != issue_number: + reasons.append( + f"durable lock targets issue #{lock.get('issue_number')}, not " + f"#{issue_number}; refusing head refresh" + ) + + for field, expected in (("remote", remote), ("org", org), ("repo", repo)): + actual = str(lock.get(field) or "").strip() + if actual != str(expected or "").strip(): + reasons.append( + f"lock {field} '{actual}' does not match requested " + f"'{str(expected or '').strip()}'" + ) + + locked_branch = str(lock.get("branch_name") or "").strip() + if locked_branch != str(branch_name or "").strip(): + reasons.append( + f"lock branch '{locked_branch}' does not match requested " + f"'{str(branch_name or '').strip()}'" + ) + + locked_worktree = str(lock.get("worktree_path") or "").strip() + try: + same_wt = bool(locked_worktree) and bool(worktree_path) and ( + os.path.realpath(locked_worktree) == os.path.realpath(worktree_path) + ) + except OSError: + same_wt = locked_worktree == (worktree_path or "") + if not same_wt: + reasons.append( + f"lock worktree '{locked_worktree}' does not match declared " + f"'{str(worktree_path or '').strip()}'" + ) + + claimant = _lock_claimant_view(lock) + locked_identity = str(claimant.get("username") or "").strip() + locked_profile = str(claimant.get("profile") or "").strip() + if not locked_identity or not locked_profile: + reasons.append( + "durable lock does not record a claimant identity/profile; " + "ownership could not be proven for head refresh" + ) + if not str(identity or "").strip() or not str(profile or "").strip(): + reasons.append( + "active session identity/profile is unknown; ownership could not be " + "proven for head refresh" + ) + if locked_identity and str(identity or "").strip() and locked_identity != str(identity).strip(): + reasons.append( + f"lock claimant '{locked_identity}' does not match active identity " + f"'{str(identity).strip()}'" + ) + if locked_profile and str(profile or "").strip() and locked_profile != str(profile).strip(): + reasons.append( + f"lock profile '{locked_profile}' does not match active profile " + f"'{str(profile).strip()}'" + ) + + # The refresh is written by the LIVE owning author session. A refresh is not + # a recovery: the current process must be the recorded owner. + recorded_pid = lock.get("session_pid") + if recorded_pid is None: + recorded_pid = lock.get("pid") + evidence["recorded_pid"] = recorded_pid + evidence["current_pid"] = current_pid + if current_pid is None: + reasons.append("current session pid is unknown; cannot prove live ownership") + else: + try: + if recorded_pid is None or int(recorded_pid) != int(current_pid): + reasons.append( + f"durable lock is owned by pid {recorded_pid}, not the current " + f"session pid {current_pid}; head refresh requires the live owner" + ) + except (TypeError, ValueError): + reasons.append( + "durable lock owner pid is malformed; cannot prove live ownership" + ) + + if not old: + reasons.append("expected_old_head is not a full 40-char hex SHA (fail closed)") + if not new: + reasons.append("new_head is not a full 40-char hex SHA (fail closed)") + if old and new and old == new: + reasons.append( + "new head equals the expected old head; a sync must advance the head" + ) + + # Compare-and-swap on the recorded head: if the lock already records a synced + # head it must be exactly the expected old head, else another sync moved it. + recorded_synced = _norm_sha(lock.get("synced_pr_head")) + evidence["recorded_synced_pr_head"] = recorded_synced + if recorded_synced is not None and old is not None and recorded_synced != old: + reasons.append( + f"durable lock already records synced head {recorded_synced}, not the " + f"expected old head {old}; a concurrent sync changed it (CAS fail closed)" + ) + + if reasons: + return {"allowed": False, "reasons": reasons, "evidence": evidence, + "expected_generation": lock_generation(lock)} + + return { + "allowed": True, + "reasons": [ + f"durable lock for issue #{issue_number} branch '{locked_branch}' is " + f"owned by the live session; refresh recorded head {old} -> {new}" + ], + "evidence": evidence, + "expected_generation": lock_generation(lock), + } + + +def apply_durable_lock_head_refresh( + *, + remote: str, + org: str, + repo: str, + issue_number: int, + branch_name: str, + worktree_path: str, + pr_number: int | None, + identity: str | None, + profile: str | None, + current_pid: int | None, + expected_old_head: str | None, + new_head: str | None, + synced_at: str, + base_head: str | None = None, + provenance: str = LOCK_HEAD_REFRESH_PROVENANCE_MERGE_SYNC, + lock_dir: str | None = None, +) -> dict[str, Any]: + """CAS-refresh the durable lock's recorded head after a branch sync (#871). + + Reads the durable lock from disk, re-asserts ownership via + ``assess_durable_lock_head_refresh``, and — only when permitted — writes the + new synced head through ``bind_session_lock`` with a generation compare-and- + swap. Then re-reads the lock and proves it records the complete new head + (read-after-write). Any failure at any step returns ``refreshed=False`` with + reasons; the caller must treat that as a partial lifecycle failure and never + report a fully successful synchronization. + """ + existing = load_issue_lock( + remote=remote, org=org, repo=repo, issue_number=issue_number, lock_dir=lock_dir + ) + assessment = assess_durable_lock_head_refresh( + existing, + remote=remote, + org=org, + repo=repo, + issue_number=issue_number, + branch_name=branch_name, + worktree_path=worktree_path, + pr_number=pr_number, + identity=identity, + profile=profile, + current_pid=current_pid, + expected_old_head=expected_old_head, + new_head=new_head, + base_head=base_head, + ) + result: dict[str, Any] = { + "refreshed": False, + "read_after_write_ok": False, + "prior_head": _norm_sha(expected_old_head), + "new_head": _norm_sha(new_head), + "reasons": list(assessment.get("reasons") or []), + "evidence": assessment.get("evidence"), + } + if not assessment.get("allowed"): + return result + + new = _norm_sha(new_head) + old = _norm_sha(expected_old_head) + record = dict(existing or {}) + sync_block = { + "last_synced_pr_head": new, + "prior_pr_head": old, + "base_head": _norm_sha(base_head), + "pr_number": pr_number, + "provenance": provenance, + "synced_at": synced_at, + "synced_by_pid": current_pid, + "synced_by": { + "username": str(identity or "").strip() or None, + "profile": str(profile or "").strip() or None, + }, + } + record["synced_pr_head"] = new + record["branch_sync"] = sync_block + history = record.get("branch_sync_history") + if not isinstance(history, list): + history = [] + history = list(history) + history.append(sync_block) + record["branch_sync_history"] = history + + try: + bind_session_lock( + record, + lock_dir=lock_dir, + expected_generation=assessment.get("expected_generation"), + renewal_sanctioned=True, + ) + except Exception as exc: # CAS miss or write failure — partial lifecycle failure + result["reasons"].append( + f"durable lock head refresh write failed (fail closed): {exc}" + ) + return result + + after = load_issue_lock( + remote=remote, org=org, repo=repo, issue_number=issue_number, lock_dir=lock_dir + ) + after_head = _norm_sha((after or {}).get("synced_pr_head")) + result["lock_generation_after"] = lock_generation(after) + if after_head == new and new is not None: + result["refreshed"] = True + result["read_after_write_ok"] = True + result["reasons"].append( + f"durable lock recorded head refreshed to {new} and verified by " + "read-after-write" + ) + else: + result["reasons"].append( + "read-after-write verification failed: durable lock does not record " + f"the new head {new} (found {after_head}); partial lifecycle failure" + ) + return result \ No newline at end of file diff --git a/issue_lock_worktree.py b/issue_lock_worktree.py index de52ff7..8188536 100644 --- a/issue_lock_worktree.py +++ b/issue_lock_worktree.py @@ -145,6 +145,175 @@ def read_head_ancestry( return result +def read_merge_sync_provenance( + worktree_path: str, + *, + prior_head_sha: str | None, + synced_head_sha: str | None, +) -> dict: + """Observe whether ``synced_head_sha`` is a sanctioned merge-based branch sync + that advanced the PR branch past ``prior_head_sha`` (#871). + + ``gitea_update_pr_branch_by_merge`` advances a PR branch by merging the base + branch *into* the branch (``POST /pulls/{n}/update?style=merge``). The result + is a merge commit ``M`` on the branch whose **first** parent is the prior + branch head and whose second parent is the base tip. When the owning session + then dies without the durable lock's recorded head being refreshed, the local + worktree still sits at ``prior_head_sha`` while the live PR head is ``M``. + + Recovering that drift safely requires proving the remote head is *exactly* + such a merge-sync — not a rewrite, rebase, force-push, or an unrelated + commit. This is that server-side observation. It reports facts only; the + disposition lives in ``issue_lock_recovery``. Every field is read from git in + the declared worktree — nothing is supplied by, or reachable from, an MCP + caller (#871). + + Provenance is proven only when ALL hold: + + * both commits are present (a rewritten/force-moved prior head leaves the + object graph and fails closed); + * ``prior_head_sha`` is a strict ancestor of ``synced_head_sha`` (the branch + history is preserved, never replaced); + * ``synced_head_sha`` is a merge commit (two or more parents), i.e. a base + merged in — a plain fast-forward of new direct commits is not a sync; + * ``prior_head_sha`` is an ancestor of the merge's **first** parent, so the + branch mainline (first-parent lineage) still reaches the prior head — a + rebase/force-push that re-authored the branch side fails this. + """ + path = (worktree_path or "").strip() + prior = (prior_head_sha or "").strip() + synced = (synced_head_sha or "").strip() + result: dict = { + "prior_head_sha": prior or None, + "synced_head_sha": synced or None, + "probe_ok": False, + "prior_present": False, + "synced_present": False, + "prior_is_ancestor": False, + "synced_is_merge": False, + "first_parent_reaches_prior": False, + "is_merge_sync": False, + "first_parent_sha": None, + "parent_count": None, + "proof": None, + "reasons": [], + } + if not path or not prior or not synced: + result["reasons"].append( + "merge-sync provenance probe requires a worktree path and both " + "commit SHAs" + ) + return result + if prior == synced: + result["reasons"].append( + "prior and synced heads are identical; no branch sync occurred" + ) + return result + + def _present(sha: str) -> bool: + res = subprocess.run( + ["git", "-C", path, "rev-parse", "--verify", "--quiet", f"{sha}^{{commit}}"], + capture_output=True, + text=True, + check=False, + ) + return res.returncode == 0 + + def _is_ancestor(ancestor: str, descendant: str) -> bool | None: + res = subprocess.run( + ["git", "-C", path, "merge-base", "--is-ancestor", ancestor, descendant], + capture_output=True, + text=True, + check=False, + ) + if res.returncode == 0: + return True + if res.returncode == 1: + return False + return None # failed probe — never a silent "no" + + try: + result["prior_present"] = _present(prior) + result["synced_present"] = _present(synced) + except OSError as exc: # git unavailable — fail closed, never assume + result["reasons"].append(f"merge-sync provenance probe could not run: {exc}") + return result + + if not result["prior_present"]: + result["reasons"].append( + f"prior head {prior} is not reachable in '{path}'; history may have " + "been rewritten or force-moved" + ) + if not result["synced_present"]: + result["reasons"].append( + f"synced head {synced} is not reachable in '{path}'" + ) + if not (result["prior_present"] and result["synced_present"]): + return result + + ancestor = _is_ancestor(prior, synced) + if ancestor is None: + result["reasons"].append( + "ancestry probe failed; merge-sync provenance unproven" + ) + return result + result["prior_is_ancestor"] = bool(ancestor) + if not ancestor: + result["reasons"].append( + f"prior head {prior} is not an ancestor of synced head {synced}; " + "the branch history was not preserved (not a merge-based sync)" + ) + return result + + parents_res = subprocess.run( + ["git", "-C", path, "rev-list", "--parents", "-n", "1", synced], + capture_output=True, + text=True, + check=False, + ) + if parents_res.returncode != 0: + result["reasons"].append( + f"could not read parents of {synced}; merge-sync provenance unproven" + ) + return result + tokens = (parents_res.stdout or "").split() + # tokens[0] is the commit itself; the rest are its parents. + parents = tokens[1:] + result["parent_count"] = len(parents) + result["synced_is_merge"] = len(parents) >= 2 + if not result["synced_is_merge"]: + result["probe_ok"] = True + result["reasons"].append( + f"synced head {synced} has {len(parents)} parent(s); a merge-based " + "branch sync produces a merge commit (two or more parents)" + ) + return result + first_parent = parents[0] + result["first_parent_sha"] = first_parent + + fp_reaches = _is_ancestor(prior, first_parent) if prior != first_parent else True + if fp_reaches is None: + result["reasons"].append( + "first-parent ancestry probe failed; merge-sync provenance unproven" + ) + return result + result["first_parent_reaches_prior"] = bool(fp_reaches) + result["probe_ok"] = True + if not fp_reaches: + result["reasons"].append( + f"merge first parent {first_parent} does not reach prior head " + f"{prior}; the branch mainline was re-authored (not a sanctioned sync)" + ) + return result + + result["is_merge_sync"] = True + result["proof"] = ( + f"synced head {synced} is a merge commit (parents={len(parents)}) whose " + f"first-parent lineage reaches prior head {prior}; base merged into branch" + ) + return result + + def read_recorded_base( worktree_path: str, *, diff --git a/mcp_restart_paths.py b/mcp_restart_paths.py new file mode 100644 index 0000000..087fe72 --- /dev/null +++ b/mcp_restart_paths.py @@ -0,0 +1,475 @@ +"""Inventory and fail-closed guards for MCP restart/reload/kill paths (#657). + +Single source of truth enumerating every code/script/doc path that can +restart, reload, reconnect, kill, or force-recreate an MCP process. Each path +is classified and linked to the guard that constrains it. The companion +human-readable inventory lives in ``docs/mcp-restart-path-inventory.md`` and is +kept in lock-step with this module by ``tests/test_mcp_restart_paths.py``. + +Design intent (aligns with #655 restart-coordinator roadmap): + +* **No unguarded full restart.** The in-process MCP daemon + (``gitea_mcp_server.py`` / ``mcp_server.py`` / ``role_session_router.py``) + must never replace or kill its own process — replacing the process after the + host wired up the stdio pipes desyncs the JSON-RPC transport (observed with + Antigravity/Cascade hosts). ``assert_no_daemon_self_replacement`` enforces + this against the live source tree. +* **No legacy auto-restart helper.** ``_trigger_mcp_auto_restart`` was removed + when the stale-runtime resolver became side-effect free (#685); + ``assert_auto_restart_helper_absent`` keeps it removed. +* **Unknown restart attempts fail closed.** LLM tools must route any restart + intent through a *registered* path. ``assert_restart_attempt_registered`` + raises ``UnknownRestartPathError`` for anything not in this inventory. +* **pkill stays forbidden (#630).** Manual daemon kills are classified as + contamination by :mod:`runtime_recovery_guard`; this module records that path + and the test asserts the classification still holds. + +This module performs no restarts, spawns no threads, and touches no config or +process state. It is pure inventory + read-only source assertions. +""" + +from __future__ import annotations + +import os +from dataclasses import dataclass +from pathlib import Path +from typing import Iterable + +# --- Classifications ------------------------------------------------------- + +#: A narrow, one-shot recovery that is safe by construction (e.g. a CLI wrapper +#: re-execing into the venv interpreter before importing anything, or an +#: in-process profile switch). Never targets the running MCP daemon process. +CLASS_SANCTIONED_NARROW = "sanctioned_narrow_recovery" + +#: The path detects a condition that would require a restart, then *fails +#: closed* on mutations and emits restart/reconnect guidance. It never restarts +#: the process itself (recovery is owned by the host/operator). +CLASS_GUARDED_FAIL_CLOSED = "guarded_fail_closed" + +#: The path is forbidden. Attempting it is a workflow-safety violation and, +#: where an LLM tool could invoke it, is marked as contamination. +CLASS_FORBIDDEN = "forbidden" + +#: A previously-existing unguarded restart primitive that has been deleted. A +#: regression guard keeps it absent. +CLASS_REMOVED = "removed" + +#: Behavior that lives in the host/IDE and is outside this process's control +#: (e.g. a manual ``/mcp reconnect``). Documented, not code-guarded here. +CLASS_HOST_RESIDUAL = "host_residual" + +VALID_CLASSIFICATIONS = frozenset( + { + CLASS_SANCTIONED_NARROW, + CLASS_GUARDED_FAIL_CLOSED, + CLASS_FORBIDDEN, + CLASS_REMOVED, + CLASS_HOST_RESIDUAL, + } +) + +#: The in-process MCP daemon modules. These must never self-replace/self-kill. +DAEMON_MODULES = ( + "gitea_mcp_server.py", + "mcp_server.py", + "role_session_router.py", +) + +#: The legacy auto-restart helper removed in #685. Must stay removed. +LEGACY_AUTO_RESTART_HELPER = "_trigger_mcp_auto_restart" + +#: Call patterns that would let the daemon replace or terminate its own +#: process. Matched as calls (trailing ``(``) so prose/docstring mentions such +#: as "we do NOT os.execv() here" or "never calls ``os._exit``" do not trip the +#: scanner (comment lines are stripped first regardless). +DAEMON_SELF_REPLACEMENT_PRIMITIVES = ( + "os.execv(", + "os.execve(", + "os.execvp(", + "os.execvpe(", + "os.kill(", + "os.killpg(", + "os._exit(", + "os.abort(", +) + + +@dataclass(frozen=True) +class RestartPath: + """One classified restart/reload/kill path in the inventory.""" + + path_id: str + title: str + mechanism: str + classification: str + guard: str + locations: tuple[str, ...] + references: tuple[str, ...] + residual_host: bool = False + notes: str = "" + + +class UnknownRestartPathError(RuntimeError): + """Raised when a restart attempt is not a registered, classified path.""" + + +# --- The inventory --------------------------------------------------------- + +_RESTART_PATHS: tuple[RestartPath, ...] = ( + RestartPath( + path_id="cli_venv_bootstrap_execv", + title="CLI wrapper venv re-exec", + mechanism=( + "Standalone CLI scripts re-exec into venv/bin/python3 via os.execv " + "at import top, guarded by `sys.executable != venv_python`." + ), + classification=CLASS_SANCTIONED_NARROW, + guard=( + "One-shot, pre-import bootstrap; runs before any MCP transport " + "exists and only when not already on the venv interpreter, so it " + "cannot desync a live daemon. Idempotent guard condition prevents " + "a re-exec loop." + ), + locations=( + "create_pr.py", + "create_issue.py", + "close_issue.py", + "merge_pr.py", + "review_pr.py", + "edit_pr.py", + "delete_branch.py", + "mark_issue.py", + "manage_labels.py", + "list_issues.py", + "list_prs.py", + ), + references=("#657",), + ), + RestartPath( + path_id="daemon_self_replacement", + title="MCP daemon self-replacement", + mechanism=( + "The in-process MCP daemon replacing/terminating its own process " + "(os.execv/os.kill/os._exit) to reload code." + ), + classification=CLASS_FORBIDDEN, + guard=( + "Forbidden by design: replacing the process after the host wired " + "up stdio desyncs JSON-RPC (Antigravity/Cascade). Enforced against " + "the source tree by assert_no_daemon_self_replacement()." + ), + locations=("gitea_mcp_server.py:~155 (decision comment)",) + DAEMON_MODULES, + references=("#657", "#584"), + ), + RestartPath( + path_id="legacy_auto_restart_helper", + title="Legacy _trigger_mcp_auto_restart helper", + mechanism=( + "A helper that actively restarted the MCP server from the " + "read-only resolver path." + ), + classification=CLASS_REMOVED, + guard=( + "Removed in #685 when the resolver became side-effect free. Kept " + "absent by assert_auto_restart_helper_absent()." + ), + locations=("gitea_mcp_server.py", "mcp_server.py"), + references=("#685", "#657"), + ), + RestartPath( + path_id="config_touch_reload", + title="MCP client config-touch reload", + mechanism=( + "Touching (utime) the MCP client config file to make the host " + "reload/recreate the server process." + ), + classification=CLASS_REMOVED, + guard=( + "Removed from the resolver in #685: stale-runtime detection is " + "report-only and never mutates client config, spawns threads, or " + "calls os._exit." + ), + locations=("gitea_mcp_server.py (resolve_task_capability)",), + references=("#685", "#657"), + ), + RestartPath( + path_id="master_advance_auto_restart", + title="Master-advance staleness gate", + mechanism=( + "On-disk master advancing past the running code. The master-parity " + "gate detects it and fails mutations closed with restart guidance." + ), + classification=CLASS_GUARDED_FAIL_CLOSED, + guard=( + "Detect + fail closed only; the process never self-restarts. " + "master_parity_gate captures startup parity and blocks mutations " + "while stale, emitting restart/reconnect guidance." + ), + locations=( + "master_parity_gate.py", + "gitea_mcp_server.py (gitea_assess_master_parity)", + ), + references=("#420", "#591", "#657"), + ), + RestartPath( + path_id="stale_runtime_resolver_reconnect", + title="Stale-runtime resolver reconnect guidance", + mechanism=( + "The capability resolver detecting a stale serving process and " + "reporting restart_required/stop_required for a client reconnect." + ), + classification=CLASS_GUARDED_FAIL_CLOSED, + guard=( + "Report-only (#685): returns restart_required/stop_required and an " + "exact_safe_next_action pointing at IDE/client reconnect; performs " + "no restart, thread spawn, config touch, or os._exit." + ), + locations=("gitea_mcp_server.py (gitea_resolve_task_capability)",), + references=("#685", "#657"), + ), + RestartPath( + path_id="manual_daemon_kill", + title="Manual daemon kill (pkill/killall/kill)", + mechanism=( + "Shell kills of the MCP daemon: `pkill -f mcp_server.py`, " + "`killall`, broad `pkill -f python` sweeps, or `kill ` of a " + "daemon pid." + ), + classification=CLASS_FORBIDDEN, + guard=( + "Forbidden (#630): runtime_recovery_guard classifies these as " + "contamination and gitea_record_daemon_process_kill_attempt writes " + "a durable marker that fails subsequent mutations closed. Operator " + "maintenance authorization is read only from the environment, not " + "from a tool argument." + ), + locations=( + "runtime_recovery_guard.py", + "gitea_mcp_server.py (gitea_record_daemon_process_kill_attempt)", + ), + references=("#630", "#657"), + ), + RestartPath( + path_id="conflict_marker_infra_stop", + title="Startup conflict-marker infra stop", + mechanism=( + "The daemon entrypoint scans for unresolved merge-conflict markers " + "at startup and stops (sys.exit(1)) if found." + ), + classification=CLASS_GUARDED_FAIL_CLOSED, + guard=( + "Fail-closed startup stop, not a restart: the process exits and " + "waits for the operator to resolve conflicts and relaunch. Never " + "self-restarts or loops." + ), + locations=("mcp_server.py (check_conflict_markers)",), + references=("#657",), + ), + RestartPath( + path_id="ide_client_reconnect", + title="Host/IDE MCP reconnect", + mechanism=( + "A manual `/mcp reconnect` (or equivalent host action) that the " + "IDE performs to recreate the MCP client connection." + ), + classification=CLASS_HOST_RESIDUAL, + guard=( + "Outside this process's control. It is the sanctioned recovery the " + "gates point operators toward; documented as residual host " + "behavior. No in-process code initiates it." + ), + locations=("host/IDE",), + references=("#584", "#656", "#657"), + residual_host=True, + ), + RestartPath( + path_id="profile_switch_runtime", + title="Runtime profile switch", + mechanism=( + "Switching the active execution profile at runtime " + "(dynamic-profile mode)." + ), + classification=CLASS_SANCTIONED_NARROW, + guard=( + "In-process and restart-free: runtime_switching_supported is true, " + "so a profile switch rebinds capability without recreating the " + "process. No restart primitive is invoked." + ), + locations=("gitea_mcp_server.py (gitea_activate_profile)",), + references=("#656", "#657"), + ), +) + +_BY_ID: dict[str, RestartPath] = {p.path_id: p for p in _RESTART_PATHS} + + +# --- Read-only accessors --------------------------------------------------- + + +def iter_restart_paths() -> tuple[RestartPath, ...]: + """Return the full inventory as an immutable tuple.""" + + return _RESTART_PATHS + + +def restart_path_ids() -> frozenset[str]: + """Return the set of registered path ids.""" + + return frozenset(_BY_ID) + + +def get_restart_path(path_id: str) -> RestartPath: + """Return the registered path, or raise :class:`UnknownRestartPathError`.""" + + try: + return _BY_ID[path_id] + except KeyError as exc: + raise UnknownRestartPathError( + f"unknown restart path id {path_id!r}; not in the #657 inventory" + ) from exc + + +def paths_by_classification(classification: str) -> tuple[RestartPath, ...]: + """Return all registered paths with the given classification.""" + + if classification not in VALID_CLASSIFICATIONS: + raise ValueError(f"unknown classification {classification!r}") + return tuple(p for p in _RESTART_PATHS if p.classification == classification) + + +def assert_restart_attempt_registered(path_id: str) -> RestartPath: + """Fail closed unless ``path_id`` is a registered, classified restart path. + + LLM tools that intend to trigger any restart/reload/reconnect must name a + registered path so an unknown/novel restart primitive cannot slip through + silently. Forbidden and removed paths are registered too — this only + asserts the attempt is *known*, not that it is *permitted*; callers must + still honor the classification. + """ + + return get_restart_path(path_id) + + +def assert_registry_wellformed() -> None: + """Validate the inventory's own invariants (fail closed on drift).""" + + seen: set[str] = set() + for path in _RESTART_PATHS: + if path.path_id in seen: + raise ValueError(f"duplicate restart path id {path.path_id!r}") + seen.add(path.path_id) + if path.classification not in VALID_CLASSIFICATIONS: + raise ValueError( + f"{path.path_id!r} has invalid classification " + f"{path.classification!r}" + ) + if not path.guard.strip(): + raise ValueError(f"{path.path_id!r} is missing a guard description") + if not path.references: + raise ValueError(f"{path.path_id!r} is missing references") + if not path.locations: + raise ValueError(f"{path.path_id!r} is missing locations") + if path.classification == CLASS_HOST_RESIDUAL and not path.residual_host: + raise ValueError( + f"{path.path_id!r} is host_residual but residual_host is False" + ) + + +# --- Source-tree guards ---------------------------------------------------- + + +def _repo_root(root: str | os.PathLike[str] | None = None) -> Path: + if root is not None: + return Path(root) + return Path(__file__).resolve().parent + + +def _iter_code_lines(text: str) -> Iterable[tuple[int, str]]: + """Yield (1-based lineno, line) for lines that are not full-line comments.""" + + for lineno, line in enumerate(text.splitlines(), start=1): + if line.lstrip().startswith("#"): + continue + yield lineno, line + + +def scan_daemon_self_replacement( + root: str | os.PathLike[str] | None = None, +) -> list[dict[str, object]]: + """Return violations where a daemon module could self-replace/self-kill. + + Scans :data:`DAEMON_MODULES` for calls in + :data:`DAEMON_SELF_REPLACEMENT_PRIMITIVES`. Full-line comments are ignored, + and only call forms (with a trailing ``(``) match, so decision comments and + docstrings that merely mention the primitives do not produce false hits. + """ + + repo = _repo_root(root) + violations: list[dict[str, object]] = [] + for module in DAEMON_MODULES: + path = repo / module + if not path.exists(): + continue + text = path.read_text(encoding="utf-8", errors="replace") + for lineno, line in _iter_code_lines(text): + for primitive in DAEMON_SELF_REPLACEMENT_PRIMITIVES: + if primitive in line: + violations.append( + { + "module": module, + "line": lineno, + "primitive": primitive, + "text": line.strip(), + } + ) + return violations + + +def assert_no_daemon_self_replacement( + root: str | os.PathLike[str] | None = None, +) -> None: + """Fail closed if any daemon module can restart/kill its own process.""" + + violations = scan_daemon_self_replacement(root) + if violations: + rendered = "; ".join( + f"{v['module']}:{v['line']} {v['primitive']}" for v in violations + ) + raise AssertionError( + "MCP daemon must never self-replace/self-kill (#657); found: " + f"{rendered}" + ) + + +def scan_auto_restart_helper( + root: str | os.PathLike[str] | None = None, +) -> list[dict[str, object]]: + """Return occurrences of a *definition* of the legacy auto-restart helper.""" + + repo = _repo_root(root) + needle = f"def {LEGACY_AUTO_RESTART_HELPER}" + hits: list[dict[str, object]] = [] + for module in DAEMON_MODULES: + path = repo / module + if not path.exists(): + continue + text = path.read_text(encoding="utf-8", errors="replace") + for lineno, line in _iter_code_lines(text): + if needle in line: + hits.append({"module": module, "line": lineno}) + return hits + + +def assert_auto_restart_helper_absent( + root: str | os.PathLike[str] | None = None, +) -> None: + """Fail closed if the removed ``_trigger_mcp_auto_restart`` reappears.""" + + hits = scan_auto_restart_helper(root) + if hits: + rendered = "; ".join(f"{h['module']}:{h['line']}" for h in hits) + raise AssertionError( + f"{LEGACY_AUTO_RESTART_HELPER} was removed in #685 and must not " + f"return (#657); found definition at: {rendered}" + ) diff --git a/merged_cleanup_reconcile.py b/merged_cleanup_reconcile.py index 4a7299a..b436750 100644 --- a/merged_cleanup_reconcile.py +++ b/merged_cleanup_reconcile.py @@ -566,6 +566,10 @@ def build_pr_cleanup_entry( worktree_state=worktree_state, active_lock=active_lock, ) + planned = plan_cleanup_execution_order( + remote_assessment=remote, + local_assessment=local, + ) return { "pr_number": pr_number, "issue_number": issue_number, @@ -576,9 +580,63 @@ def build_pr_cleanup_entry( "merged": merged, "remote_branch": remote, "local_worktree": local, + # #851: dry-run and execute share the same lifecycle order description. + "planned_execution_order": planned, } +def plan_cleanup_execution_order( + *, + remote_assessment: dict[str, Any] | None, + local_assessment: dict[str, Any] | None, +) -> list[dict[str, Any]]: + """Describe independent worktree-then-reassess-then-remote cleanup order (#851). + + Remote ownership protection remains fail-closed at execute time. A worktree + that is independently safe to remove is never skipped merely because remote + deletion may be blocked by that same ``worktree_binding``. + """ + remote = remote_assessment or {} + local = local_assessment or {} + steps: list[dict[str, Any]] = [] + worktree_safe = bool(local.get("safe_to_remove_worktree")) + remote_safe = bool(remote.get("safe_to_delete_remote")) + + if worktree_safe: + steps.append( + { + "action": "remove_local_worktree", + "reason": "independently_safe_to_remove", + "phase": 1, + } + ) + if remote_safe: + if worktree_safe: + steps.append( + { + "action": "reassess_branch_ownership", + "reason": "after_worktree_removal_clear_worktree_binding", + "phase": 2, + } + ) + steps.append( + { + "action": "delete_remote_branch", + "reason": "only_if_independently_safe_after_reassessment", + "phase": 3, + } + ) + else: + steps.append( + { + "action": "delete_remote_branch", + "reason": "safe_to_delete_and_no_independent_worktree_removal", + "phase": 1, + } + ) + return steps + + def build_reconciliation_report( *, project_root: str, diff --git a/pr_work_lease.py b/pr_work_lease.py index 2fb6876..b877f10 100644 --- a/pr_work_lease.py +++ b/pr_work_lease.py @@ -228,25 +228,74 @@ def find_active_reviewer_lease( return None +def _conflict_fix_chain_key(lease: dict) -> tuple | None: + """Identity of the lease chain a conflict-fix marker belongs to (#842). + + Keyed by PR number, profile, head_before, and branch. Returns None when any + required component (pr_number, profile, head_before) is missing or malformed. + """ + raw = lease.get("raw_fields") or {} + pr_number = lease.get("pr_number") + profile = (lease.get("profile") or "").strip().lower() + head_before = lease.get("head_before") + branch = (lease.get("branch") or raw.get("branch") or "").strip() + if not (pr_number and profile and head_before): + return None + return (pr_number, profile, head_before, branch) + + +def _conflict_fix_chain_matches(key1: tuple, key2: tuple) -> bool: + """True when two conflict-fix chain keys refer to the same lease chain.""" + pr1, profile1, head1, branch1 = key1 + pr2, profile2, head2, branch2 = key2 + if pr1 != pr2 or profile1 != profile2 or head1 != head2: + return False + if branch1 and branch2 and branch1 != branch2: + return False + return True + + +def _conflict_fix_chain_terminated_after(entries: list[dict], index: int) -> bool: + """True when a later marker terminates the conflict-fix chain of ``entries[index]``. + + Append-only newest-wins: a terminal marker (phase=released/blocked/done) + ends only its matching claim chain (#842). + """ + key = _conflict_fix_chain_key(entries[index]) + if key is None: + return False + for later in entries[index + 1:]: + phase = (later.get("phase") or "").strip().lower() + if phase not in _TERMINAL_CONFLICT_FIX_PHASES: + continue + later_key = _conflict_fix_chain_key(later) + if later_key and _conflict_fix_chain_matches(key, later_key): + return True + return False + + def find_active_conflict_fix_lease( comments: list[dict], *, pr_number: int, now: datetime | None = None, ) -> dict[str, Any] | None: - """Return the newest unexpired conflict-fix lease for *pr_number*, if any.""" + """Return the newest unexpired, non-terminated conflict-fix lease for *pr_number*, if any.""" now = now or datetime.now(timezone.utc) candidates = [ entry for entry in _comment_entries(comments, pr_number=pr_number) if entry.get("lease_kind") == "conflict_fix" ] - for lease in reversed(candidates): + for index in range(len(candidates) - 1, -1, -1): + lease = candidates[index] if _lease_expired(lease, now=now): continue phase = (lease.get("phase") or "").strip().lower() if phase in _TERMINAL_CONFLICT_FIX_PHASES: continue if phase in _ACTIVE_CONFLICT_FIX_PHASES or phase: + if _conflict_fix_chain_terminated_after(candidates, index): + continue return lease return None diff --git a/restart_coordinator.py b/restart_coordinator.py new file mode 100644 index 0000000..5db7de0 --- /dev/null +++ b/restart_coordinator.py @@ -0,0 +1,451 @@ +"""MCP restart coordinator and impact analysis (#658). + +Before any sanctioned MCP restart, a central coordinator must evaluate the +live control-plane state — active sessions, leases/locks, in-flight issue/PR +work, mutations, worktrees, and recovery history — and produce an *impact +preview* so operators (and the web console, #642/#652) can see the blast +radius **before** concurrent LLM work is disrupted. + +Design rules (mirrors the read-only posture of ``workflow_dashboard`` / +``lease_lifecycle``): + +* **Pure classification.** :func:`evaluate_restart_impact` takes an already + gathered inventory and returns a structured report. It never touches the + network, the filesystem, or a live process, so multi-session fixtures can + drive every branch in unit tests. The coordinator *never restarts anything*; + a mutative apply path is a later child gated by a drain proof (non-goal here). +* **Fail closed.** If the inventory is not explicitly complete, the verdict is + ``unsafe`` / deny — an incomplete evaluation must never green-light a restart. +* **No secrets.** Session ids, pids, and profiles are operational metadata, not + credentials; nothing secret flows through this module. + +The single sanctioned entry point post-#657 is the MCP tool +``gitea_request_mcp_restart`` (dry-run by default), which gathers the inventory +from the #613 control-plane DB and calls :func:`evaluate_restart_impact`. +""" + +from __future__ import annotations + +from dataclasses import dataclass, field +from datetime import datetime, timezone +from typing import Any, Mapping, Sequence + +import lease_lifecycle + +COORDINATOR_VERSION = "1.0.0-issue-658" + +# Restart verdicts. Exactly the three the acceptance criteria name. +VERDICT_SAFE = "safe" +VERDICT_UNSAFE = "unsafe" +VERDICT_OVERRIDE = "override" + +# Blast-radius severity bands. +BLAST_NONE = "none" +BLAST_LOW = "low" +BLAST_MEDIUM = "medium" +BLAST_HIGH = "high" + +# A live lease with a live owner process is treated as active in-flight work. +LEASE_FRESHNESS_LIVE = "active" + +# Default staleness window for a session heartbeat (seconds). A session whose +# last heartbeat is older than this is not counted as live even if its row is +# still marked ``active`` — it is assumed dead/detached. +DEFAULT_SESSION_HEARTBEAT_STALE_SECONDS = 900 + + +def _utc_now() -> datetime: + return datetime.now(timezone.utc) + + +def _parse_ts(value: str | None) -> datetime | None: + return lease_lifecycle._parse_ts(value) + + +@dataclass(frozen=True) +class SessionImpact: + """One MCP session a restart would terminate.""" + + session_id: str + role: str | None + profile: str | None + pid: int | None + status: str | None + alive: bool | None + heartbeat_stale: bool + is_requester: bool + live: bool + + def as_dict(self) -> dict[str, Any]: + return { + "session_id": self.session_id, + "role": self.role, + "profile": self.profile, + "pid": self.pid, + "status": self.status, + "alive": self.alive, + "heartbeat_stale": self.heartbeat_stale, + "is_requester": self.is_requester, + "live": self.live, + } + + +@dataclass(frozen=True) +class LeaseImpact: + """One control-plane lease a restart would disrupt.""" + + lease_id: str | None + session_id: str | None + role: str | None + phase: str | None + freshness: str | None + work_kind: str | None + work_number: int | None + worktree_path: str | None + disruptive: bool + is_mutation: bool + is_critical_section: bool + + def as_dict(self) -> dict[str, Any]: + return { + "lease_id": self.lease_id, + "session_id": self.session_id, + "role": self.role, + "phase": self.phase, + "freshness": self.freshness, + "work_kind": self.work_kind, + "work_number": self.work_number, + "worktree_path": self.worktree_path, + "disruptive": self.disruptive, + "is_mutation": self.is_mutation, + "is_critical_section": self.is_critical_section, + } + + +@dataclass(frozen=True) +class RestartImpactReport: + """Impact preview DTO returned to the console / operator (#642/#652).""" + + coordinator_version: str + evaluated_at: str + dry_run: bool + restart_performed: bool + inventory_complete: bool + verdict: str + allow_restart: bool + override_would_allow: bool + operator_override: bool + blast_radius: str + reasons: list[str] + affected_sessions: list[SessionImpact] + affected_leases: list[LeaseImpact] + critical_sections: list[LeaseImpact] + affected_issues: list[int] + affected_prs: list[int] + mutations: list[LeaseImpact] + terminal_lock: dict[str, Any] | None + ack_state: dict[str, str] + prior_recovery_attempts: list[dict[str, Any]] + counts: dict[str, int] + audit_record: dict[str, Any] + incomplete_reasons: list[str] = field(default_factory=list) + + def as_dict(self) -> dict[str, Any]: + return { + "coordinator_version": self.coordinator_version, + "evaluated_at": self.evaluated_at, + "dry_run": self.dry_run, + "restart_performed": self.restart_performed, + "inventory_complete": self.inventory_complete, + "incomplete_reasons": list(self.incomplete_reasons), + "verdict": self.verdict, + "allow_restart": self.allow_restart, + "override_would_allow": self.override_would_allow, + "operator_override": self.operator_override, + "blast_radius": self.blast_radius, + "reasons": list(self.reasons), + "affected_sessions": [s.as_dict() for s in self.affected_sessions], + "affected_leases": [l.as_dict() for l in self.affected_leases], + "critical_sections": [l.as_dict() for l in self.critical_sections], + "affected_issues": list(self.affected_issues), + "affected_prs": list(self.affected_prs), + "mutations": [l.as_dict() for l in self.mutations], + "terminal_lock": self.terminal_lock, + "ack_state": dict(self.ack_state), + "prior_recovery_attempts": list(self.prior_recovery_attempts), + "counts": dict(self.counts), + "audit_record": dict(self.audit_record), + } + + +def _classify_session( + row: Mapping[str, Any], + *, + now: datetime, + requesting_session_id: str | None, + heartbeat_stale_seconds: int, +) -> SessionImpact: + session_id = str(row.get("session_id") or "") + pid = row.get("pid") + status = (row.get("status") or "").strip().lower() or None + alive = lease_lifecycle.is_process_alive(pid) if pid is not None else None + hb = _parse_ts(row.get("last_heartbeat_at")) + heartbeat_stale = bool( + hb is not None and (now - hb).total_seconds() > heartbeat_stale_seconds + ) + live = bool(status == "active" and alive is not False and not heartbeat_stale) + return SessionImpact( + session_id=session_id, + role=row.get("role"), + profile=row.get("profile"), + pid=pid, + status=status, + alive=alive, + heartbeat_stale=heartbeat_stale, + is_requester=bool( + requesting_session_id and session_id == requesting_session_id + ), + live=live, + ) + + +# Lease phases that represent an active mutation in flight (as opposed to a +# mere allocation/claim with no work committed yet). An active lease in any of +# these phases is a critical section a restart must not sever. +_MUTATING_PHASES = frozenset( + { + "implementing", + "publishing", + "pushing", + "committing", + "reviewing", + "merging", + "reconciling", + "conflict_fix", + } +) + + +def _classify_lease(row: Mapping[str, Any]) -> LeaseImpact: + freshness_obj = row.get("freshness") + if isinstance(freshness_obj, Mapping): + freshness = str(freshness_obj.get("freshness") or "").strip().lower() or None + else: + freshness = str(freshness_obj or "").strip().lower() or None + phase = (row.get("phase") or "").strip().lower() or None + worktree = row.get("worktree_path") + disruptive = freshness == LEASE_FRESHNESS_LIVE + # A live lease is a mutation-in-flight if it carries an author worktree or + # its phase names a mutating step. All disruptive leases are critical + # sections a restart would sever regardless. + is_mutation = bool( + disruptive and (bool(worktree) or (phase in _MUTATING_PHASES)) + ) + number = row.get("work_number") + try: + number = int(number) if number is not None else None + except (TypeError, ValueError): + number = None + return LeaseImpact( + lease_id=row.get("lease_id"), + session_id=row.get("session_id"), + role=row.get("role"), + phase=phase, + freshness=freshness, + work_kind=(str(row.get("work_kind") or "").strip().lower() or None), + work_number=number, + worktree_path=worktree, + disruptive=disruptive, + is_mutation=is_mutation, + is_critical_section=disruptive, + ) + + +def _blast_radius(*, session_count: int, work_count: int, mutation_count: int) -> str: + if mutation_count > 0 or work_count >= 3 or session_count >= 3: + return BLAST_HIGH + if work_count > 0 or session_count == 2: + return BLAST_MEDIUM + if session_count == 1: + return BLAST_LOW + return BLAST_NONE + + +def evaluate_restart_impact( + inventory: Mapping[str, Any], + *, + now: datetime | None = None, + operator_override: bool = False, + requesting_session_id: str | None = None, + dry_run: bool = True, + session_heartbeat_stale_seconds: int = DEFAULT_SESSION_HEARTBEAT_STALE_SECONDS, +) -> RestartImpactReport: + """Evaluate a proposed MCP restart and return an impact preview. + + ``inventory`` is a mapping with: + + * ``sessions`` — session rows (session_id, role, profile, pid, status, + last_heartbeat_at). + * ``leases`` — control-plane lease rows, each ideally carrying an enriched + ``freshness`` dict (as :func:`lease_lifecycle.list_active_leases` returns); + a bare string freshness is also accepted. + * ``terminal_lock`` — the active terminal (merge) lock, if any. + * ``prior_recovery_attempts`` — narrower recovery attempts already tried + (e.g. sanctioned reconnects) so the operator sees escalation history. + * ``inventory_complete`` — bool. **Must** be explicitly True; a missing or + falsy value forces a deny (fail closed). + * ``incomplete_reasons`` — optional reasons the inventory is incomplete. + + The coordinator never restarts anything: ``restart_performed`` is always + False and the mutative apply path is a later drain-gated child. + """ + moment = now or _utc_now() + reasons: list[str] = [] + + inventory_complete = bool(inventory.get("inventory_complete", False)) + incomplete_reasons = [str(r) for r in (inventory.get("incomplete_reasons") or [])] + + sessions_raw: Sequence[Mapping[str, Any]] = inventory.get("sessions") or [] + leases_raw: Sequence[Mapping[str, Any]] = inventory.get("leases") or [] + terminal_lock = inventory.get("terminal_lock") or None + prior_recovery_attempts = [ + dict(a) for a in (inventory.get("prior_recovery_attempts") or []) + ] + + session_impacts = [ + _classify_session( + s, + now=moment, + requesting_session_id=requesting_session_id, + heartbeat_stale_seconds=session_heartbeat_stale_seconds, + ) + for s in sessions_raw + ] + lease_impacts = [_classify_lease(l) for l in leases_raw] + + # Only *other* live sessions and live leases constitute blast radius: a + # restart that would kill only the requesting session with no other work in + # flight is safe. + other_live_sessions = [ + s for s in session_impacts if s.live and not s.is_requester + ] + disruptive_leases = [l for l in lease_impacts if l.disruptive] + critical_sections = [l for l in lease_impacts if l.is_critical_section] + mutations = [l for l in lease_impacts if l.is_mutation] + + affected_issues = sorted( + { + l.work_number + for l in disruptive_leases + if l.work_kind == "issue" and l.work_number is not None + } + ) + affected_prs = sorted( + { + l.work_number + for l in disruptive_leases + if l.work_kind == "pr" and l.work_number is not None + } + ) + + disruptive = bool(disruptive_leases or other_live_sessions or terminal_lock) + + if not inventory_complete: + verdict = VERDICT_UNSAFE + allow_restart = False + reasons.append( + "inventory incomplete: restart evaluation cannot confirm blast " + "radius — deny (fail closed, #658)" + ) + reasons.extend(incomplete_reasons) + elif not disruptive: + verdict = VERDICT_SAFE + allow_restart = True + reasons.append("no other live sessions, live leases, or terminal lock") + elif operator_override: + verdict = VERDICT_OVERRIDE + allow_restart = True + reasons.append( + "live work present; operator override accepts the blast radius" + ) + else: + verdict = VERDICT_UNSAFE + allow_restart = False + reasons.append( + "live work would be disrupted; restart denied without operator " + "override" + ) + + if critical_sections and inventory_complete: + reasons.append( + f"{len(critical_sections)} critical section(s) in flight " + "(active lease with a live owner)" + ) + if terminal_lock: + reasons.append("active terminal (merge) lock present") + + override_would_allow = bool(inventory_complete and disruptive) + + blast_radius = _blast_radius( + session_count=len(other_live_sessions), + work_count=len(affected_issues) + len(affected_prs), + mutation_count=len(mutations), + ) + + # Acknowledgement is a later child (drain protocol); expose per-session + # placeholders so the console can render the ack column now. + ack_state = {s.session_id: "pending" for s in other_live_sessions} + + counts = { + "sessions_total": len(session_impacts), + "sessions_live_other": len(other_live_sessions), + "leases_total": len(lease_impacts), + "leases_disruptive": len(disruptive_leases), + "critical_sections": len(critical_sections), + "mutations": len(mutations), + "affected_issues": len(affected_issues), + "affected_prs": len(affected_prs), + "prior_recovery_attempts": len(prior_recovery_attempts), + } + + audit_record = { + "event": "restart_impact_evaluated", + "coordinator_version": COORDINATOR_VERSION, + "evaluated_at": moment.isoformat(), + "dry_run": dry_run, + "operator_override": bool(operator_override), + "requesting_session_id": requesting_session_id, + "inventory_complete": inventory_complete, + "verdict": verdict, + "allow_restart": allow_restart, + "blast_radius": blast_radius, + "counts": counts, + } + + return RestartImpactReport( + coordinator_version=COORDINATOR_VERSION, + evaluated_at=moment.isoformat(), + dry_run=dry_run, + restart_performed=False, + inventory_complete=inventory_complete, + verdict=verdict, + allow_restart=allow_restart, + override_would_allow=override_would_allow, + operator_override=bool(operator_override), + blast_radius=blast_radius, + reasons=reasons, + affected_sessions=session_impacts, + affected_leases=lease_impacts, + critical_sections=critical_sections, + affected_issues=affected_issues, + affected_prs=affected_prs, + mutations=mutations, + terminal_lock=dict(terminal_lock) + if isinstance(terminal_lock, Mapping) + else terminal_lock, + ack_state=ack_state, + prior_recovery_attempts=prior_recovery_attempts, + counts=counts, + audit_record=audit_record, + incomplete_reasons=incomplete_reasons, + ) diff --git a/scripts/worktree-start b/scripts/worktree-start index a189164..a79e908 100755 --- a/scripts/worktree-start +++ b/scripts/worktree-start @@ -43,19 +43,21 @@ repo_root="$(cd "$script_dir/.." && pwd)" # Enforce issue-linked, traceable branch names (issue → branch → worktree → PR). if [[ "$allow_unlinked" -eq 0 ]]; then - locked_branch=$(python3 -c " + if [[ "$dry_run" -eq 0 ]] && [[ ! "$branch" =~ ^review/pr-[0-9]+-.+ ]]; then + locked_branch=$(python3 -c " import sys sys.path.insert(0, '$repo_root') import issue_lock_store print(issue_lock_store.resolve_locked_branch_for_session('$branch')) ") - if [[ -z "$locked_branch" ]]; then - echo "Error: No session issue lock is bound. Call gitea_lock_issue before branch creation (fail closed)." >&2 - exit 2 - fi - if [[ "$branch" != "$locked_branch" ]]; then - echo "Error: Requested branch '$branch' does not match locked branch '$locked_branch' (fail closed)." >&2 - exit 2 + if [[ -z "$locked_branch" ]]; then + echo "Error: No session issue lock is bound. Call gitea_lock_issue before branch creation (fail closed)." >&2 + exit 2 + fi + if [[ "$branch" != "$locked_branch" ]]; then + echo "Error: Requested branch '$branch' does not match locked branch '$locked_branch' (fail closed)." >&2 + exit 2 + fi fi if [[ "$branch" =~ ^(fix|feat|docs|chore)/issue-[0-9]+-.+ ]] \ diff --git a/task_capability_map.py b/task_capability_map.py index 0b8ac0b..213e701 100644 --- a/task_capability_map.py +++ b/task_capability_map.py @@ -32,6 +32,26 @@ TASK_CAPABILITY_MAP: dict[str, dict[str, str]] = { "permission": "gitea.issue.comment", "role": "author", }, + # #860: dirty orphaned same-claimant worktree recovery (explicit operation). + "recover_dirty_orphaned_issue_worktree": { + "permission": "gitea.issue.comment", + "role": "author", + }, + "gitea_recover_dirty_orphaned_issue_worktree": { + "permission": "gitea.issue.comment", + "role": "author", + }, + # #864: dirty-preserving same-claimant author-session rebind (dead owner PID). + # Author MCP tool path. Reconciler execute is gated inside the tool via + # authorize_reconciler_execute + role_kind checks (not this map entry). + "rebind_dirty_same_claimant_author_session": { + "permission": "gitea.issue.comment", + "role": "author", + }, + "gitea_rebind_dirty_same_claimant_author_session": { + "permission": "gitea.issue.comment", + "role": "author", + }, "set_issue_labels": { "permission": "gitea.issue.comment", "role": "author", @@ -477,6 +497,11 @@ TASK_CAPABILITY_MAP: dict[str, dict[str, str]] = { # merger lease (#763). _PREFLIGHT_TASK_TRANSITIONS = frozenset({ ("review_pr", "acquire_reviewer_pr_lease"), + ("work_issue", "lock_issue"), + ("work_issue", "recover_dirty_orphaned_issue_worktree"), + ("work_issue", "gitea_recover_dirty_orphaned_issue_worktree"), + ("work_issue", "commit_files"), + ("work_issue", "gitea_commit_files"), }) diff --git a/tests/test_branch_cleanup_guard.py b/tests/test_branch_cleanup_guard.py index 7a78998..fddb534 100644 --- a/tests/test_branch_cleanup_guard.py +++ b/tests/test_branch_cleanup_guard.py @@ -1266,6 +1266,730 @@ class TestSecondRemediationIntegration(unittest.TestCase): self.assertIn("delete_acknowledged", delete_actions[0]) self.assertTrue(delete_actions[0].get("verified_absent")) + def test_issue_851_worktree_removed_when_remote_blocked_only_by_worktree_binding(self): + """#851: remote blocked by worktree_binding must not skip safe worktree removal. + + Lifecycle: remove clean owned worktree → reassess ownership → delete + remote only if independently safe. Unrelated entries stay untouched. + """ + from mcp_server import gitea_reconcile_merged_cleanups + + target_branch = "fix/issue-844-exclude-epic-containers" + foreign_branch = "fix/issue-999-unrelated-active" + worktree_path = "/tmp/branches/fix-issue-844-exclude-epic-containers" + ownership_calls = [] + remove_calls = [] + delete_api_calls = [] + + def fake_collect(**kwargs): + ownership_calls.append(dict(kwargs)) + # Ownership is reassessed *after* independent worktree removal (#851). + # Target worktree is already gone → no worktree_binding remains. + # Foreign branch keeps an active author lease → remote delete blocked. + if kwargs.get("branch") == foreign_branch: + # Match session-bound org/repo + host used by the tool resolve path. + return { + "records": [ + { + "category": guard.OWNERSHIP_CATEGORY_AUTHOR_LEASE, + "status": "active", + "remote": kwargs.get("remote") or "prgs", + "host": kwargs.get("host") or "gitea.example.com", + "org": kwargs.get("org") or "Scaled-Tech-Consulting", + "repo": kwargs.get("repo") or "Gitea-Tools", + "branch": foreign_branch, + "reclaim_allowed": False, + } + ], + "inventory_error": False, + } + return {"records": [], "inventory_error": False} + + def fake_remove(project_root, branch, worktree_path=None): + remove_calls.append( + {"branch": branch, "worktree_path": worktree_path} + ) + return { + "success": True, + "performed": True, + "message": f"removed worktree {worktree_path}", + "worktree_path": worktree_path, + } + + def fake_probe(h, o, r, auth, br): + return guard.classify_branch_readback_http_status( + 404, not_found_scope=guard.NOT_FOUND_SCOPE_BRANCH + ) + + def fake_api(method, url, auth, **kwargs): + if method == "DELETE": + delete_api_calls.append(url) + return {} + + report = { + "entries": [ + { + "pr_number": 848, + "head_branch": target_branch, + "remote_branch": {"safe_to_delete_remote": True}, + "local_worktree": { + "safe_to_remove_worktree": True, + "worktree_path": worktree_path, + }, + }, + { + "pr_number": 999, + "head_branch": foreign_branch, + "remote_branch": {"safe_to_delete_remote": True}, + "local_worktree": { + "safe_to_remove_worktree": False, + "worktree_path": None, + }, + }, + ], + "reviewer_scratch_entries": [], + } + patch( + "mcp_server.get_profile", + return_value={ + "profile_name": "prgs-reconciler", + "role": "reconciler", + "allowed_operations": [ + "gitea.read", + "gitea.branch.delete", + "gitea.pr.close", + ], + "forbidden_operations": [], + }, + ).start() + patch("mcp_server.api_get_all", return_value=[]).start() + patch( + "mcp_server.merged_cleanup_reconcile.build_reconciliation_report", + return_value=report, + ).start() + patch( + "mcp_server.merged_cleanup_reconcile.discover_reviewer_scratch_worktrees", + return_value=[], + ).start() + patch( + "mcp_server.audit_reconciliation_mode.check_cleanup_execution_allowed", + return_value=(True, []), + ).start() + patch("mcp_server.verify_preflight_purity", return_value=None).start() + patch( + "mcp_server._collect_branch_ownership_records", + side_effect=fake_collect, + ).start() + patch("mcp_server._probe_remote_branch", side_effect=fake_probe).start() + patch( + "mcp_server.merged_cleanup_reconcile.remove_local_worktree", + side_effect=fake_remove, + ).start() + self.mock_api.side_effect = fake_api + + res = gitea_reconcile_merged_cleanups( + dry_run=False, + execute_confirmed=True, + remote="prgs", + ) + self.assertTrue(res.get("performed") or res.get("executed")) + actions = res.get("actions") or [] + + remove_actions = [ + a for a in actions if a.get("action") == "remove_local_worktree" + ] + self.assertEqual(len(remove_actions), 1, actions) + self.assertTrue(remove_actions[0].get("success")) + self.assertEqual(remove_calls[0]["branch"], target_branch) + self.assertEqual(remove_calls[0]["worktree_path"], worktree_path) + + # Target remote delete succeeds after worktree removal + reassessment. + target_deletes = [ + a + for a in actions + if a.get("action") == "delete_remote_branch" + and a.get("branch") == target_branch + ] + self.assertEqual(len(target_deletes), 1, actions) + self.assertTrue(target_deletes[0].get("success")) + self.assertTrue(target_deletes[0].get("after_worktree_removal")) + self.assertTrue(target_deletes[0].get("ownership_reassessed")) + self.assertTrue(target_deletes[0].get("verified_absent")) + + # Foreign branch remains protected (author lease) and is not deleted. + foreign_deletes = [ + a + for a in actions + if a.get("action") == "delete_remote_branch" + and a.get("branch") == foreign_branch + ] + self.assertEqual(len(foreign_deletes), 1, actions) + self.assertFalse(foreign_deletes[0].get("success")) + self.assertEqual( + foreign_deletes[0].get("blocker_kind"), "active_branch_ownership" + ) + self.assertIn( + guard.OWNERSHIP_CATEGORY_AUTHOR_LEASE, + foreign_deletes[0].get("blocking_categories") or [], + ) + # Only the target branch should hit the DELETE API. + self.assertEqual(len(delete_api_calls), 1) + + # Ownership collected for target (post-removal) and foreign; worktree + # removal happened before target remote delete in the action log. + target_idx = next( + i + for i, a in enumerate(actions) + if a.get("action") == "remove_local_worktree" + ) + delete_idx = next( + i + for i, a in enumerate(actions) + if a.get("action") == "delete_remote_branch" + and a.get("branch") == target_branch + and a.get("success") + ) + self.assertLess(target_idx, delete_idx) + + def test_issue_851_dirty_worktree_not_removed_and_remote_stays_protected(self): + """#851: dirty/foreign worktrees remain protected; no unsafe cleanup.""" + from mcp_server import gitea_reconcile_merged_cleanups + + branch = "fix/issue-851-dirty" + remove_calls = [] + + def fake_collect(**kwargs): + return { + "records": [ + { + "category": guard.OWNERSHIP_CATEGORY_WORKTREE_BINDING, + "status": "active", + "remote": kwargs.get("remote") or "prgs", + "host": kwargs.get("host") or "gitea.example.com", + "org": kwargs.get("org") or "Scaled-Tech-Consulting", + "repo": kwargs.get("repo") or "Gitea-Tools", + "branch": branch, + "reclaim_allowed": False, + } + ], + "inventory_error": False, + } + + report = { + "entries": [ + { + "pr_number": 851, + "head_branch": branch, + "remote_branch": {"safe_to_delete_remote": True}, + "local_worktree": { + "safe_to_remove_worktree": False, + "worktree_path": "/tmp/dirty-wt", + }, + } + ], + "reviewer_scratch_entries": [], + } + patch( + "mcp_server.get_profile", + return_value={ + "profile_name": "prgs-reconciler", + "role": "reconciler", + "allowed_operations": [ + "gitea.read", + "gitea.branch.delete", + ], + "forbidden_operations": [], + }, + ).start() + patch("mcp_server.api_get_all", return_value=[]).start() + patch( + "mcp_server.merged_cleanup_reconcile.build_reconciliation_report", + return_value=report, + ).start() + patch( + "mcp_server.merged_cleanup_reconcile.discover_reviewer_scratch_worktrees", + return_value=[], + ).start() + patch( + "mcp_server.audit_reconciliation_mode.check_cleanup_execution_allowed", + return_value=(True, []), + ).start() + patch("mcp_server.verify_preflight_purity", return_value=None).start() + patch( + "mcp_server._collect_branch_ownership_records", + side_effect=fake_collect, + ).start() + patch( + "mcp_server.merged_cleanup_reconcile.remove_local_worktree", + side_effect=lambda *a, **k: remove_calls.append(k) or { + "success": True, + "performed": True, + }, + ).start() + self.mock_api.side_effect = lambda *a, **k: {} + + res = gitea_reconcile_merged_cleanups( + dry_run=False, + execute_confirmed=True, + remote="prgs", + ) + actions = res.get("actions") or [] + self.assertEqual(remove_calls, []) + self.assertFalse( + any(a.get("action") == "remove_local_worktree" for a in actions) + ) + deletes = [ + a for a in actions if a.get("action") == "delete_remote_branch" + ] + self.assertEqual(len(deletes), 1) + self.assertFalse(deletes[0].get("success")) + self.assertEqual(deletes[0].get("blocker_kind"), "active_branch_ownership") + self.assertIn( + guard.OWNERSHIP_CATEGORY_WORKTREE_BINDING, + deletes[0].get("blocking_categories") or [], + ) + + def test_issue_851_idempotent_resume_when_worktree_already_absent(self): + """#851: partial failures remain resumable and idempotent.""" + from mcp_server import gitea_reconcile_merged_cleanups + + branch = "fix/issue-851-resume" + ownership_calls = [] + + def fake_collect(**kwargs): + ownership_calls.append(kwargs) + return {"records": [], "inventory_error": False} + + def fake_remove(project_root, branch, worktree_path=None): + return { + "success": False, + "performed": False, + "message": f"worktree not found: {worktree_path}", + } + + def fake_probe(h, o, r, auth, br): + return guard.classify_branch_readback_http_status( + 404, not_found_scope=guard.NOT_FOUND_SCOPE_BRANCH + ) + + report = { + "entries": [ + { + "pr_number": 851, + "head_branch": branch, + "remote_branch": {"safe_to_delete_remote": True}, + "local_worktree": { + "safe_to_remove_worktree": True, + "worktree_path": "/tmp/already-gone", + }, + } + ], + "reviewer_scratch_entries": [], + } + patch( + "mcp_server.get_profile", + return_value={ + "profile_name": "prgs-reconciler", + "role": "reconciler", + "allowed_operations": [ + "gitea.read", + "gitea.branch.delete", + ], + "forbidden_operations": [], + }, + ).start() + patch("mcp_server.api_get_all", return_value=[]).start() + patch( + "mcp_server.merged_cleanup_reconcile.build_reconciliation_report", + return_value=report, + ).start() + patch( + "mcp_server.merged_cleanup_reconcile.discover_reviewer_scratch_worktrees", + return_value=[], + ).start() + patch( + "mcp_server.audit_reconciliation_mode.check_cleanup_execution_allowed", + return_value=(True, []), + ).start() + patch("mcp_server.verify_preflight_purity", return_value=None).start() + patch( + "mcp_server._collect_branch_ownership_records", + side_effect=fake_collect, + ).start() + patch("mcp_server._probe_remote_branch", side_effect=fake_probe).start() + patch( + "mcp_server.merged_cleanup_reconcile.remove_local_worktree", + side_effect=fake_remove, + ).start() + self.mock_api.side_effect = lambda *a, **k: {} + + res = gitea_reconcile_merged_cleanups( + dry_run=False, + execute_confirmed=True, + remote="prgs", + ) + actions = res.get("actions") or [] + removes = [a for a in actions if a.get("action") == "remove_local_worktree"] + deletes = [a for a in actions if a.get("action") == "delete_remote_branch"] + self.assertEqual(len(removes), 1) + self.assertFalse(removes[0].get("success")) + self.assertEqual(len(deletes), 1) + self.assertTrue(deletes[0].get("success")) + self.assertTrue(deletes[0].get("after_worktree_removal")) + self.assertTrue(ownership_calls) + + +class TestIssue855ExactPrSelector(unittest.TestCase): + """#855: exact pr_number pin for reconcile_merged_cleanups (#851 lifecycle).""" + + def setUp(self): + self._remotes = patch.dict( + mcp_server.REMOTES, + { + "prgs": { + "host": "gitea.example.com", + "org": "Scaled-Tech-Consulting", + "repo": "Gitea-Tools", + } + }, + ) + self._remotes.start() + patch("gitea_audit.audit_enabled", return_value=False).start() + self.mock_api = patch("mcp_server.api_request").start() + self.mock_all = patch("mcp_server.api_get_all", return_value=[]).start() + patch("mcp_server.get_auth_header", return_value=FAKE_AUTH).start() + patch( + "mcp_server.merged_cleanup_reconcile.is_head_ancestor_of_ref", + return_value=True, + ).start() + patch( + "mcp_server.get_profile", + return_value=dict(RECONCILER_WITH_DELETE), + ).start() + patch( + "mcp_server._profile_operation_gate", + return_value=[], + ).start() + patch( + "mcp_server._collect_branch_ownership_records", + return_value={"records": [], "inventory_error": False}, + ).start() + patch( + "mcp_server.merged_cleanup_reconcile.discover_reviewer_scratch_worktrees", + return_value=[], + ).start() + patch("mcp_server.verify_preflight_purity", return_value=None).start() + patch( + "mcp_server.audit_reconciliation_mode.check_cleanup_execution_allowed", + return_value=(True, []), + ).start() + + def tearDown(self): + patch.stopall() + + def _merged_pr(self, number, branch, sha="c" * 40): + return { + "number": number, + "title": f"PR {number}", + "body": f"Closes #{number - 4}", + "merged": True, + "merged_at": "2026-07-23T12:00:00Z", + "merge_commit_sha": "f" * 40, + "state": "closed", + "head": {"ref": branch, "sha": sha}, + "base": {"ref": "master"}, + } + + def test_exact_pr_848_ignores_newer_852_in_batch_queue(self): + """pr_number=848 selects only #848 even when #852 is newer/first.""" + from mcp_server import gitea_reconcile_merged_cleanups + + pr_848 = self._merged_pr( + 848, "fix/issue-844-exclude-epic-containers", sha="c3f282ba" + "0" * 32 + ) + # Closed list would rank #852 first in batch mode; exact pin must ignore it. + closed_batch = [ + self._merged_pr(852, "fix/issue-851-cleanup-worktree-before-remote-delete"), + pr_848, + self._merged_pr(849, "fix/issue-849-other"), + self._merged_pr(846, "fix/issue-846-other"), + self._merged_pr(845, "fix/issue-845-other"), + ] + batch_fetch_calls = [] + + def fake_api(method, url, *args, **kwargs): + if method == "GET" and url.rstrip("/").endswith("/pulls/848"): + return dict(pr_848) + if method == "GET" and "/pulls/" in url: + raise AssertionError(f"unexpected PR fetch: {url}") + if method == "GET" and "/branches/" in url: + return {"name": "present"} + return {} + + def fake_all(url, auth, limit=None): + batch_fetch_calls.append((url, limit)) + if "state=open" in url: + return [] + if "state=closed" in url: + # Exact mode must not use the closed batch list. + raise AssertionError( + "exact pr_number mode must not page closed PRs: " + url + ) + return [] + + self.mock_api.side_effect = fake_api + self.mock_all.side_effect = fake_all + patch( + "mcp_server._remote_branch_exists", + return_value=True, + ).start() + patch( + "mcp_server.merged_cleanup_reconcile.build_reconciliation_report", + side_effect=lambda **kwargs: { + "entries": [ + { + "pr_number": int(pr["number"]), + "head_branch": (pr.get("head") or {}).get("ref"), + "issue_number": 844, + "remote_branch": { + "safe_to_delete_remote": True, + "head_branch": (pr.get("head") or {}).get("ref"), + }, + "local_worktree": { + "safe_to_remove_worktree": True, + "worktree_path": ( + "/tmp/branches/fix-issue-844-exclude-epic-containers" + ), + }, + "planned_execution_order": ( + mcp_server.merged_cleanup_reconcile.plan_cleanup_execution_order( + remote_assessment={"safe_to_delete_remote": True}, + local_assessment={"safe_to_remove_worktree": True}, + ) + ), + } + for pr in kwargs.get("closed_prs") or [] + if pr.get("merged_at") or pr.get("merged") + ], + "reviewer_scratch_entries": [], + "merged_pr_count": len(kwargs.get("closed_prs") or []), + }, + ).start() + + res = gitea_reconcile_merged_cleanups( + dry_run=True, + pr_number=848, + remote="prgs", + org="Scaled-Tech-Consulting", + repo="Gitea-Tools", + ) + self.assertTrue(res.get("success")) + self.assertFalse(res.get("performed")) + self.assertEqual(res.get("selection_mode"), "exact_pr") + self.assertEqual(res.get("selected_pr_number"), 848) + entries = res.get("entries") or [] + self.assertEqual(len(entries), 1, entries) + self.assertEqual(entries[0].get("pr_number"), 848) + self.assertEqual( + entries[0].get("head_branch"), + "fix/issue-844-exclude-epic-containers", + ) + # No other PR appears in plan. + self.assertEqual(list((res.get("planned_execution_orders") or {}).keys()), ["848"]) + plan = (res.get("planned_execution_orders") or {}).get("848") or [] + actions = [s.get("action") for s in plan] + self.assertEqual( + actions, + [ + "remove_local_worktree", + "reassess_branch_ownership", + "delete_remote_branch", + ], + ) + # Prove we never scanned the multi-PR closed batch. + self.assertFalse(any("state=closed" in (u or "") for u, _ in batch_fetch_calls)) + # closed_batch fixture must remain unused (sanity). + self.assertEqual(closed_batch[0]["number"], 852) + + def test_exact_pr_execute_only_mutates_selected_pr(self): + """Execute with pr_number must never touch #845/#846/#849/#852.""" + from mcp_server import gitea_reconcile_merged_cleanups + + pr_848 = self._merged_pr(848, "fix/issue-844-exclude-epic-containers") + worktree_path = "/tmp/branches/fix-issue-844-exclude-epic-containers" + remove_calls = [] + delete_api_calls = [] + ownership_branches = [] + + def fake_api(method, url, *args, **kwargs): + if method == "GET" and url.rstrip("/").endswith("/pulls/848"): + return dict(pr_848) + if method == "DELETE": + delete_api_calls.append(url) + # Forbid foreign PR branch deletion by URL content. + for forbidden in ("845", "846", "849", "852"): + self.assertNotIn(forbidden, url) + return {} + + def fake_remove(project_root, branch, worktree_path=None): + remove_calls.append({"branch": branch, "worktree_path": worktree_path}) + return { + "success": True, + "performed": True, + "message": f"removed {worktree_path}", + "worktree_path": worktree_path, + } + + def fake_collect(**kwargs): + ownership_branches.append(kwargs.get("branch")) + return {"records": [], "inventory_error": False} + + def fake_probe(h, o, r, auth, br): + return guard.classify_branch_readback_http_status( + 404, not_found_scope=guard.NOT_FOUND_SCOPE_BRANCH + ) + + self.mock_api.side_effect = fake_api + self.mock_all.side_effect = lambda url, auth, limit=None: [] + patch("mcp_server._remote_branch_exists", return_value=True).start() + patch( + "mcp_server.merged_cleanup_reconcile.build_reconciliation_report", + return_value={ + "entries": [ + { + "pr_number": 848, + "head_branch": "fix/issue-844-exclude-epic-containers", + "remote_branch": {"safe_to_delete_remote": True}, + "local_worktree": { + "safe_to_remove_worktree": True, + "worktree_path": worktree_path, + }, + "planned_execution_order": [ + {"action": "remove_local_worktree", "phase": 1}, + {"action": "reassess_branch_ownership", "phase": 2}, + {"action": "delete_remote_branch", "phase": 3}, + ], + } + ], + "reviewer_scratch_entries": [ + # Foreign scratch must be filtered before report execute loop; + # if present here it would still be a test failure if acted on. + ], + "merged_pr_count": 1, + }, + ).start() + patch( + "mcp_server.merged_cleanup_reconcile.remove_local_worktree", + side_effect=fake_remove, + ).start() + patch( + "mcp_server._collect_branch_ownership_records", + side_effect=fake_collect, + ).start() + patch("mcp_server._probe_remote_branch", side_effect=fake_probe).start() + + res = gitea_reconcile_merged_cleanups( + dry_run=False, + execute_confirmed=True, + pr_number=848, + remote="prgs", + org="Scaled-Tech-Consulting", + repo="Gitea-Tools", + ) + self.assertTrue(res.get("performed") or res.get("executed")) + self.assertEqual(res.get("selection_mode"), "exact_pr") + self.assertEqual(res.get("selected_pr_number"), 848) + actions = res.get("actions") or [] + pr_numbers_touched = { + a.get("pr_number") for a in actions if a.get("pr_number") is not None + } + self.assertTrue(pr_numbers_touched.issubset({None, 848}) or not pr_numbers_touched) + removes = [a for a in actions if a.get("action") == "remove_local_worktree"] + deletes = [a for a in actions if a.get("action") == "delete_remote_branch"] + self.assertEqual(len(removes), 1) + self.assertEqual(remove_calls[0]["branch"], "fix/issue-844-exclude-epic-containers") + self.assertEqual(len(deletes), 1) + self.assertTrue(deletes[0].get("success")) + self.assertTrue(deletes[0].get("after_worktree_removal")) + self.assertEqual(len(delete_api_calls), 1) + self.assertEqual( + ownership_branches, ["fix/issue-844-exclude-epic-containers"] + ) + + def test_exact_pr_unknown_fails_closed_without_mutation(self): + from mcp_server import gitea_reconcile_merged_cleanups + + def fake_api(method, url, *args, **kwargs): + if method == "GET" and "/pulls/99999" in url: + raise RuntimeError("HTTP 404 Not Found") + raise AssertionError(f"unexpected API call {method} {url}") + + self.mock_api.side_effect = fake_api + res = gitea_reconcile_merged_cleanups( + dry_run=True, + pr_number=99999, + remote="prgs", + ) + self.assertFalse(res.get("success")) + self.assertFalse(res.get("performed")) + self.assertEqual(res.get("blocker_kind"), "pr_unresolvable") + self.assertIn("99999", " ".join(res.get("reasons") or [])) + + def test_exact_pr_not_merged_fails_closed(self): + from mcp_server import gitea_reconcile_merged_cleanups + + def fake_api(method, url, *args, **kwargs): + if method == "GET" and url.rstrip("/").endswith("/pulls/900"): + return { + "number": 900, + "merged": False, + "merged_at": None, + "state": "open", + "head": {"ref": "feat/x", "sha": "a" * 40}, + } + raise AssertionError(f"unexpected {method} {url}") + + self.mock_api.side_effect = fake_api + res = gitea_reconcile_merged_cleanups( + dry_run=False, + execute_confirmed=True, + pr_number=900, + remote="prgs", + ) + self.assertFalse(res.get("success")) + self.assertFalse(res.get("performed")) + self.assertEqual(res.get("blocker_kind"), "pr_not_merged") + + def test_exact_pr_invalid_number_fails_closed(self): + from mcp_server import gitea_reconcile_merged_cleanups + + res = gitea_reconcile_merged_cleanups( + dry_run=True, + pr_number=0, + remote="prgs", + ) + self.assertFalse(res.get("success")) + self.assertEqual(res.get("blocker_kind"), "invalid_pr_number") + self.mock_api.assert_not_called() + + def test_batch_mode_still_works_without_pr_number(self): + """Unfiltered batch path remains backward compatible.""" + from mcp_server import gitea_reconcile_merged_cleanups + + self.mock_all.side_effect = lambda url, auth, limit=None: [] + self.mock_api.side_effect = lambda *a, **k: {} + patch( + "mcp_server.merged_cleanup_reconcile.build_reconciliation_report", + return_value={ + "entries": [], + "reviewer_scratch_entries": [], + "merged_pr_count": 0, + }, + ).start() + res = gitea_reconcile_merged_cleanups(dry_run=True, remote="prgs", limit=10) + self.assertTrue(res.get("success")) + self.assertEqual(res.get("selection_mode"), "batch") + self.assertIsNone(res.get("selected_pr_number")) if __name__ == "__main__": diff --git a/tests/test_dirty_orphan_worktree_recovery.py b/tests/test_dirty_orphan_worktree_recovery.py new file mode 100644 index 0000000..4e58b4a --- /dev/null +++ b/tests/test_dirty_orphan_worktree_recovery.py @@ -0,0 +1,483 @@ +"""Synthetic regression coverage for dirty orphaned worktree recovery (#860). + +Modeled on the #850 / #855 shape without mutating their real state. +""" + +from __future__ import annotations + +import json +import os +import shutil +import tempfile +import unittest +from unittest import mock + +import dirty_orphan_worktree_recovery as dorec +import issue_lock_store + + +DEAD_PID = 999_999_999 +LIVE_PID = os.getpid() +BRANCH = "fix/issue-901-dirty-orphan" +SOURCE_WT = "/repo/branches/issue-901-dirty-orphan" +RECOVERY_WT_NAME = "recovery-issue-901-dirty-orphan" +LOCAL_HEAD = "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa" +REMOTE_HEAD = "bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb" +OTHER_HEAD = "cccccccccccccccccccccccccccccccccccccccc" +FP_A = dorec.sha256_bytes(b"dirty-a") +FP_B = dorec.sha256_bytes(b"dirty-b") +FP_C = dorec.sha256_bytes(b"dirty-c-conflict") + + +def durable_lock(**overrides): + """#850-shaped PID-less malformed same-claimant lock.""" + lock = { + "issue_number": 901, + "branch_name": BRANCH, + "worktree_path": SOURCE_WT, + "remote": "prgs", + "org": "Example-Org", + "repo": "Example-Repo", + # intentionally no pid / session_pid / work_lease expiry + "claimant": {"username": "author-user", "profile": "prgs-author"}, + } + lock.update(overrides) + return lock + + +def base_kwargs(**overrides): + kwargs = { + "issue_number": 901, + "branch_name": BRANCH, + "source_worktree_path": SOURCE_WT, + "remote": "prgs", + "org": "Example-Org", + "repo": "Example-Repo", + "identity": "author-user", + "profile": "prgs-author", + "expected_local_head": LOCAL_HEAD, + "expected_remote_head": REMOTE_HEAD, + "expected_dirty_fingerprints": {"a.py": FP_A, "b.py": FP_B}, + "current_branch": BRANCH, + "porcelain_status": " M a.py\n M b.py\n", + "observed_local_head": LOCAL_HEAD, + "observed_remote_head": REMOTE_HEAD, + "observed_dirty_fingerprints": {"a.py": FP_A, "b.py": FP_B}, + "competing_live_locks": [], + "competing_live_sessions": [], + "workflow_lease_active": False, + "workflow_lease_expired": True, + "canonical_repo_root": "/repo", + "worktree_registered": True, + "current_pid": LIVE_PID, + } + kwargs.update(overrides) + return kwargs + + +def assess(lock=None, **overrides): + return dorec.assess_dirty_orphan_recovery( + durable_lock() if lock is None else lock, **base_kwargs(**overrides) + ) + + +class FreshnessPidLess(unittest.TestCase): + def test_pid_less_lock_is_not_live(self): + freshness = issue_lock_store.assess_lock_freshness(durable_lock()) + self.assertFalse(freshness["live"]) + self.assertTrue(freshness.get("pid_missing")) + self.assertEqual(freshness["status"], "malformed") + + def test_pid_less_with_far_future_expiry_still_not_live(self): + lock = durable_lock( + work_lease={ + "operation_type": "author_issue_work", + "expires_at": "2999-01-01T00:00:00Z", + "last_heartbeat_at": "2999-01-01T00:00:00Z", + } + ) + freshness = issue_lock_store.assess_lock_freshness(lock) + self.assertFalse(freshness["live"]) + self.assertTrue(freshness.get("pid_missing")) + + +class EligibilityGranted(unittest.TestCase): + def test_dead_same_claimant_pid_less_dirty(self): + result = assess() + self.assertEqual(result["outcome"], dorec.ELIGIBLE) + self.assertTrue(result["eligible"]) + + def test_expired_workflow_lease_corroboration(self): + result = assess(workflow_lease_active=False, workflow_lease_expired=True) + self.assertTrue(result["eligible"]) + + def test_older_local_newer_remote_heads(self): + result = assess() + self.assertTrue(result["evidence"].get("heads_diverged")) + self.assertTrue(result["eligible"]) + + +class EligibilityRefused(unittest.TestCase): + def test_active_owner_with_pid(self): + lock = durable_lock(pid=LIVE_PID, session_pid=LIVE_PID) + result = assess(lock=lock, owner_process_alive_override=True) + self.assertEqual(result["outcome"], dorec.REFUSED) + self.assertFalse(result["eligible"]) + self.assertTrue(any("alive" in r for r in result["reasons"])) + + def test_foreign_claimant(self): + result = assess(identity="other-user") + self.assertEqual(result["outcome"], dorec.REFUSED) + self.assertTrue(any("foreign claimant identity" in r for r in result["reasons"])) + + def test_foreign_profile(self): + result = assess(profile="prgs-reviewer") + self.assertEqual(result["outcome"], dorec.REFUSED) + + def test_fingerprint_mismatch(self): + result = assess(observed_dirty_fingerprints={"a.py": "0" * 64, "b.py": FP_B}) + self.assertEqual(result["outcome"], dorec.REFUSED) + self.assertTrue(any("fingerprint mismatch" in r for r in result["reasons"])) + + def test_head_mismatch(self): + result = assess(observed_local_head=OTHER_HEAD) + self.assertEqual(result["outcome"], dorec.REFUSED) + + def test_remote_head_mismatch(self): + result = assess(observed_remote_head=OTHER_HEAD) + self.assertEqual(result["outcome"], dorec.REFUSED) + + def test_path_not_under_branches(self): + result = assess( + source_worktree_path="/tmp/branches/evil", + # lock path also changed so worktree agreement holds + lock=durable_lock(worktree_path="/tmp/branches/evil"), + ) + self.assertEqual(result["outcome"], dorec.REFUSED) + self.assertTrue(any("canonical branches" in r for r in result["reasons"])) + + def test_unregistered_worktree(self): + result = assess(worktree_registered=False) + self.assertEqual(result["outcome"], dorec.REFUSED) + + def test_active_workflow_lease(self): + result = assess(workflow_lease_active=True, workflow_lease_expired=False) + self.assertEqual(result["outcome"], dorec.REFUSED) + + def test_unsafe_dirty_path_pin(self): + result = assess( + expected_dirty_fingerprints={"../etc/passwd": FP_A}, + observed_dirty_fingerprints={"../etc/passwd": FP_A}, + ) + self.assertEqual(result["outcome"], dorec.REFUSED) + + def test_symlink_escape_rejected_by_ancestry(self): + ok, reasons = dorec.is_path_under_canonical_branches( + "/tmp/branches/evil", canonical_repo_root="/repo" + ) + self.assertFalse(ok) + self.assertTrue(reasons) + + +class ConflictDetection(unittest.TestCase): + def test_overlapping_upstream_change(self): + conflicts = dorec.detect_path_conflicts( + dirty_paths=["c.py"], + local_head_contents={"c.py": b"local-base"}, + remote_head_contents={"c.py": b"remote-changed"}, + dirty_contents={"c.py": b"dirty-c-conflict"}, + ) + self.assertEqual(len(conflicts), 1) + self.assertEqual(conflicts[0]["path"], "c.py") + + def test_unchanged_upstream_no_conflict(self): + conflicts = dorec.detect_path_conflicts( + dirty_paths=["a.py"], + local_head_contents={"a.py": b"same"}, + remote_head_contents={"a.py": b"same"}, + dirty_contents={"a.py": b"dirty-a"}, + ) + self.assertEqual(conflicts, []) + + +class CrashSafeRecovery(unittest.TestCase): + def setUp(self): + self.tmp = tempfile.mkdtemp(prefix="dirty-orphan-") + self.repo = os.path.join(self.tmp, "repo") + self.branches = os.path.join(self.repo, "branches") + self.source = os.path.join(self.branches, "issue-901-dirty-orphan") + self.recovery = os.path.join(self.branches, RECOVERY_WT_NAME) + os.makedirs(self.source, exist_ok=True) + os.makedirs(self.branches, exist_ok=True) + # seed dirty files in source + with open(os.path.join(self.source, "a.py"), "wb") as fh: + fh.write(b"dirty-a") + with open(os.path.join(self.source, "b.py"), "wb") as fh: + fh.write(b"dirty-b") + self.journal_dir = os.path.join(self.tmp, "journals") + self.lock = durable_lock(worktree_path=self.source) + self.assessment = dorec.assess_dirty_orphan_recovery( + self.lock, + **base_kwargs( + source_worktree_path=self.source, + canonical_repo_root=self.repo, + ), + ) + + class FakeGit(dorec.GitOps): + def __init__(self, recovery_path, head): + self.recovery_path = recovery_path + self.head = head + self.calls = [] + + def run(self, args, *, cwd): + self.calls.append((args, cwd)) + if args[:3] == ["git", "worktree", "add"]: + os.makedirs(self.recovery_path, exist_ok=True) + return mock.Mock(returncode=0, stdout="", stderr="") + if args[:2] == ["git", "checkout"]: + return mock.Mock(returncode=0, stdout="", stderr="") + if args[:2] == ["git", "rev-parse"]: + return mock.Mock(returncode=0, stdout=self.head + "\n", stderr="") + return mock.Mock(returncode=0, stdout="", stderr="") + + self.git = FakeGit(self.recovery, REMOTE_HEAD) + self.written_locks = [] + + def lock_writer(record): + self.written_locks.append(record) + + self.lock_writer = lock_writer + + def tearDown(self): + shutil.rmtree(self.tmp, ignore_errors=True) + + def _run(self, **overrides): + kwargs = { + "assessment": self.assessment, + "existing_lock": self.lock, + "issue_number": 901, + "branch_name": BRANCH, + "source_worktree_path": self.source, + "recovery_worktree_path": self.recovery, + "remote": "prgs", + "org": "Example-Org", + "repo": "Example-Repo", + "identity": "author-user", + "profile": "prgs-author", + "expected_local_head": LOCAL_HEAD, + "expected_remote_head": REMOTE_HEAD, + "expected_dirty_fingerprints": {"a.py": FP_A, "b.py": FP_B}, + "dirty_contents": {"a.py": b"dirty-a", "b.py": b"dirty-b"}, + "local_head_contents": {"a.py": b"base-a", "b.py": b"base-b"}, + "remote_head_contents": {"a.py": b"base-a", "b.py": b"base-b"}, + "canonical_repo_root": self.repo, + "bind_lock": True, + "lock_writer": self.lock_writer, + "git_ops": self.git, + "journal_dir": self.journal_dir, + "session_pid": LIVE_PID, + } + kwargs.update(overrides) + return dorec.run_dirty_orphan_recovery(**kwargs) + + def test_success_preserves_dirty_bytes_and_source(self): + result = self._run() + self.assertTrue(result["success"]) + self.assertEqual(result["outcome"], dorec.RECOVERY_COMPLETED) + self.assertTrue(os.path.isdir(self.source)) + with open(os.path.join(self.source, "a.py"), "rb") as fh: + self.assertEqual(fh.read(), b"dirty-a") + with open(os.path.join(self.recovery, "a.py"), "rb") as fh: + self.assertEqual(fh.read(), b"dirty-a") + with open(os.path.join(self.recovery, "b.py"), "rb") as fh: + self.assertEqual(fh.read(), b"dirty-b") + self.assertEqual(len(self.written_locks), 1) + rec = self.written_locks[0] + self.assertEqual(rec["session_pid"], LIVE_PID) + self.assertTrue(rec["dirty_orphan_recovery"]["recovered"]) + self.assertTrue(rec["dirty_orphan_recovery"]["source_frozen"]) + + def test_conflict_leaves_governed_state(self): + result = self._run( + expected_dirty_fingerprints={"c.py": FP_C}, + dirty_contents={"c.py": b"dirty-c-conflict"}, + local_head_contents={"c.py": b"local-base"}, + remote_head_contents={"c.py": b"remote-changed"}, + ) + # #860 F4: session binding is NOT finalized while conflicts remain + self.assertFalse(result["success"]) + self.assertEqual(result["outcome"], dorec.CONFLICTS_PRESENT) + sidecar = os.path.join(self.recovery, "c.py.recovered-dirty") + self.assertTrue(os.path.isfile(sidecar)) + state = os.path.join( + self.recovery, dorec.CONFLICT_STATE_DIR, dorec.CONFLICT_STATE_FILE + ) + self.assertTrue(os.path.isfile(state)) + with open(state, "r", encoding="utf-8") as fh: + payload = json.load(fh) + self.assertEqual(payload["resolution"], "author_edit_required") + + def test_interrupt_before_journal_no_artifacts(self): + result = self._run(interrupt_after_phase=dorec.PHASE_ELIGIBILITY) + self.assertFalse(result["success"]) + self.assertEqual(result["outcome"], "INTERRUPTED") + self.assertFalse(os.path.isdir(self.recovery)) + + def test_interrupt_after_journal_then_retry_idempotent(self): + first = self._run(interrupt_after_phase=dorec.PHASE_JOURNAL_PERSISTED) + self.assertEqual(first["outcome"], "INTERRUPTED") + self.assertTrue(first["journal"]["artifacts_created"]["journal"]) + second = self._run() + self.assertTrue(second["success"]) + # source still recoverable + with open(os.path.join(self.source, "a.py"), "rb") as fh: + self.assertEqual(fh.read(), b"dirty-a") + + def test_interrupt_after_worktree_then_retry(self): + first = self._run(interrupt_after_phase=dorec.PHASE_RECOVERY_WORKTREE) + self.assertEqual(first["outcome"], "INTERRUPTED") + self.assertTrue(os.path.isdir(self.recovery)) + second = self._run() + self.assertTrue(second["success"]) + + def test_interrupt_after_binding_then_retry_complete(self): + first = self._run(interrupt_after_phase=dorec.PHASE_BINDING) + self.assertEqual(first["outcome"], "INTERRUPTED") + second = self._run() + self.assertTrue(second["success"]) + # completed journal makes further retries no-ops + third = self._run() + self.assertEqual(third["outcome"], dorec.RECOVERY_RESUMED) + + def test_source_worktree_never_deleted(self): + self._run() + self.assertTrue(os.path.isdir(self.source)) + self.assertTrue(os.path.isfile(os.path.join(self.source, "a.py"))) + + def test_fingerprint_drift_refuses_without_mutation(self): + result = self._run(dirty_contents={"a.py": b"CHANGED", "b.py": b"dirty-b"}) + self.assertFalse(result["success"]) + self.assertFalse(os.path.isdir(self.recovery)) + + +class SessionBindingPreflight(unittest.TestCase): + def test_canonical_session_binding_recognized(self): + lock = { + "worktree_path": "/repo/branches/recovery", + "session_pid": LIVE_PID, + "dirty_orphan_recovery": { + "recovered": True, + "conflicts": [], + "recovery_worktree_path": "/repo/branches/recovery", + "source_worktree_path": SOURCE_WT, + "accepted_head": REMOTE_HEAD, + }, + } + result = dorec.preflight_recognizes_recovered_provenance(lock) + self.assertTrue(result["recognized"]) + + def test_conflicts_block_commit_preflight(self): + lock = { + "worktree_path": "/repo/branches/recovery", + "session_pid": LIVE_PID, + "dirty_orphan_recovery": { + "recovered": True, + "conflicts": [{"path": "c.py"}], + }, + } + result = dorec.preflight_recognizes_recovered_provenance(lock) + self.assertFalse(result["recognized"]) + + def test_active_foreign_does_not_mutate(self): + # assess-only path: foreign refused before run + result = assess(identity="intruder") + self.assertFalse(result["eligible"]) + + +class JournalSymlinkRefusal(unittest.TestCase): + def test_symlink_journal_path_refused_on_load(self): + tmp = tempfile.mkdtemp() + try: + real = os.path.join(tmp, "real.json") + with open(real, "w", encoding="utf-8") as fh: + fh.write("{}") + link = os.path.join(tmp, "link.json") + os.symlink(real, link) + key = "symlink-test" + jdir = tmp + path = dorec._journal_path(key, journal_dir=jdir) + with open(path, "w", encoding="utf-8") as fh: + json.dump({"idempotency_key": key}, fh) + os.remove(path) + os.symlink(real, path) + with self.assertRaises(ValueError): + dorec.load_journal(key, journal_dir=jdir) + finally: + shutil.rmtree(tmp, ignore_errors=True) + + +class RealGitMultiWorktreeIntegration(unittest.TestCase): + def setUp(self): + import subprocess + self.tmp = tempfile.mkdtemp(prefix="git-integration-") + self.repo = os.path.join(self.tmp, "repo") + os.makedirs(self.repo, exist_ok=True) + subprocess.run(["git", "init"], cwd=self.repo, check=True, capture_output=True) + subprocess.run(["git", "config", "user.name", "Test User"], cwd=self.repo, check=True) + subprocess.run(["git", "config", "user.email", "test@example.com"], cwd=self.repo, check=True) + with open(os.path.join(self.repo, "init.txt"), "w") as fh: + fh.write("init") + subprocess.run(["git", "add", "."], cwd=self.repo, check=True) + subprocess.run(["git", "commit", "-m", "init"], cwd=self.repo, check=True) + branch = "fix/issue-999-test" + subprocess.run(["git", "branch", branch], cwd=self.repo, check=True) + self.branches = os.path.join(self.repo, "branches") + self.source = os.path.join(self.branches, "issue-999-test") + subprocess.run(["git", "worktree", "add", self.source, branch], cwd=self.repo, check=True) + self.dirty_path = os.path.join(self.source, "dirty.txt") + with open(self.dirty_path, "w") as fh: + fh.write("dirty-data") + + def tearDown(self): + shutil.rmtree(self.tmp, ignore_errors=True) + + def test_prepare_recovery_worktree_detached_no_exit_128(self): + import subprocess + head_sha = subprocess.check_output(["git", "rev-parse", "HEAD"], cwd=self.repo, text=True).strip() + rec_wt = os.path.join(self.branches, "recovery-issue-999-test") + res = dorec.prepare_recovery_worktree( + canonical_repo_root=self.repo, + recovery_worktree_path=rec_wt, + branch_name="fix/issue-999-test", + remote_head=head_sha, + ) + self.assertTrue(res["success"], res.get("reasons")) + self.assertTrue(os.path.isdir(rec_wt)) + + def test_real_lock_rebind_recovery_sanctioned(self): + lock_dir = os.path.join(self.tmp, "locks") + rec_wt = os.path.join(self.branches, "recovery-issue-999-test") + os.makedirs(rec_wt, exist_ok=True) + record = { + "remote": "prgs", + "org": "Example-Org", + "repo": "Example-Repo", + "issue_number": 999, + "branch_name": "fix/issue-999-test", + "worktree_path": rec_wt, + "claimant": {"username": "author-user", "profile": "prgs-author"}, + } + record_src = dict(record) + record_src["worktree_path"] = self.source + issue_lock_store.bind_session_lock(record_src, lock_dir=lock_dir) + path = issue_lock_store.bind_session_lock( + record, + lock_dir=lock_dir, + recovery_sanctioned=True, + ) + self.assertTrue(os.path.isfile(path)) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/test_dirty_same_claimant_session_rebind.py b/tests/test_dirty_same_claimant_session_rebind.py new file mode 100644 index 0000000..c40933e --- /dev/null +++ b/tests/test_dirty_same_claimant_session_rebind.py @@ -0,0 +1,1346 @@ +"""Integration tests for dirty same-claimant author-session rebind (#864 / #868). + +Uses real temp git repos/worktrees and a temp GITEA_ISSUE_LOCK_DIR. Does not +mutate any real #860/#864/#868 worktree on disk. + +#868 adds complete dirty-inventory revalidation around bind_session_lock and +complete recovery-journal operation identity (remote/org/repo/claimant). +""" + +from __future__ import annotations + +import json +import os +import subprocess +import sys +import tempfile +from datetime import datetime, timedelta, timezone +from pathlib import Path + +import pytest + +sys.path.insert(0, str(Path(__file__).resolve().parent.parent)) + +import dirty_same_claimant_session_rebind as rebind # noqa: E402 +import issue_lock_provenance # noqa: E402 +import issue_lock_store as ils # noqa: E402 +import issue_lock_worktree # noqa: E402 + +ISSUE = 864 +BRANCH = f"fix/issue-{ISSUE}-dirty-same-claimant-session-rebind" +REMOTE = "prgs" +ORG = "Scaled-Tech-Consulting" +REPO = "Gitea-Tools" +IDENTITY = "jcwalker3" +PROFILE = "prgs-author" + + +def _git(cwd: str, *args: str, check: bool = True) -> subprocess.CompletedProcess: + return subprocess.run( + ["git", "-C", cwd, *args], + capture_output=True, + text=True, + check=check, + ) + + +def dead_pid() -> int: + proc = subprocess.Popen([sys.executable, "-c", "pass"]) + proc.wait() + return proc.pid + + +def future_ts(hours: int = 4) -> str: + return ( + (datetime.now(timezone.utc) + timedelta(hours=hours)) + .isoformat() + .replace("+00:00", "Z") + ) + + +@pytest.fixture +def lock_dir(tmp_path, monkeypatch): + d = tmp_path / "issue-locks" + d.mkdir() + monkeypatch.setenv("GITEA_ISSUE_LOCK_DIR", str(d)) + return str(d) + + +@pytest.fixture +def dirty_repo(tmp_path): + """Canonical repo root with branches/ worktree and dirty content.""" + root = tmp_path / "repo" + root.mkdir() + main = root / "main" + main.mkdir() + subprocess.run(["git", "init", "-q", str(main)], check=True, capture_output=True) + _git(str(main), "config", "user.email", "t@t") + _git(str(main), "config", "user.name", "t") + (main / "README.md").write_text("base\n", encoding="utf-8") + _git(str(main), "add", "README.md") + _git(str(main), "commit", "-q", "-m", "base") + _git(str(main), "branch", "-M", "master") + + # Bare remote + origin tracking so remote head is observable offline. + bare = tmp_path / "remote.git" + subprocess.run( + ["git", "init", "--bare", "-q", str(bare)], check=True, capture_output=True + ) + _git(str(main), "remote", "add", "origin", str(bare)) + _git(str(main), "push", "-q", "origin", "master:master") + + branches = root / "branches" + branches.mkdir() + wt_name = f"fix-issue-{ISSUE}-dirty-same-claimant-session-rebind" + wt = branches / wt_name + _git(str(main), "worktree", "add", "-q", "-b", BRANCH, str(wt)) + _git(str(wt), "push", "-q", "-u", "origin", BRANCH) + + # Seed committed files we will dirty. + tracked = [ + "dirty_same_claimant_session_rebind.py", + "issue_lock_provenance.py", + "task_capability_map.py", + ] + for rel in tracked: + p = wt / rel + p.parent.mkdir(parents=True, exist_ok=True) + p.write_text(f"seed {rel}\n", encoding="utf-8") + _git(str(wt), "add", *tracked) + _git(str(wt), "commit", "-q", "-m", "seed tracked") + _git(str(wt), "push", "-q", "origin", BRANCH) + + # Dirty tracked + untracked. + for rel in tracked: + (wt / rel).write_text(f"dirty {rel}\n", encoding="utf-8") + untracked = [ + "tests/test_dirty_same_claimant_session_rebind.py", + "docs/runbook-dirty-rebind.md", + "scratch/notes-untracked.txt", + "extra_untracked.txt", + ] + for rel in untracked: + p = wt / rel + p.parent.mkdir(parents=True, exist_ok=True) + p.write_text(f"untracked {rel}\n", encoding="utf-8") + + inv = rebind.collect_dirty_inventory(str(wt)) + assert inv["ok"], inv.get("reasons") + head = _git(str(wt), "rev-parse", "HEAD").stdout.strip() + remote_head = _git( + str(wt), "rev-parse", f"refs/remotes/origin/{BRANCH}" + ).stdout.strip() + assert head == remote_head + + return { + "root": str(root), + "main": str(main), + "worktree": str(wt), + "branch": BRANCH, + "inventory": inv, + "local_head": head, + "remote_head": remote_head, + "dirty_paths": list(inv["dirty_paths"]), + "fingerprints": dict(inv["fingerprints"]), + } + + +def _make_lock( + *, + worktree: str, + pid: int, + lock_dir: str, + identity: str = IDENTITY, + profile: str = PROFILE, + **overrides, +) -> dict: + lease = { + "operation_type": ils.AUTHOR_ISSUE_WORK_LEASE, + "issue_number": ISSUE, + "branch": BRANCH, + "worktree_path": worktree, + "claimant": {"username": identity, "profile": profile}, + "created_at": "2026-01-01T00:00:00Z", + "expires_at": future_ts(), + "last_heartbeat_at": "2026-01-01T00:00:00Z", + } + lock = { + "issue_number": ISSUE, + "branch_name": BRANCH, + "worktree_path": worktree, + "remote": REMOTE, + "org": ORG, + "repo": REPO, + "session_pid": pid, + "pid": pid, + "work_lease": lease, + "lock_generation": 1, + "lock_provenance": issue_lock_provenance.build_sanctioned_lock_provenance( + tool="gitea_lock_issue", + claimant={"username": identity, "profile": profile}, + ), + } + lock.update(overrides) + path = ils.lock_file_path( + remote=REMOTE, org=ORG, repo=REPO, issue_number=ISSUE, lock_dir=lock_dir + ) + lock["lock_file_path"] = path + ils.save_lock_file(path, lock) + # Stale session pointer for the dead owner. + ptr = { + "pid": pid, + "lock_file_path": path, + "issue_number": ISSUE, + "branch_name": BRANCH, + "remote": REMOTE, + "org": ORG, + "repo": REPO, + } + ils.save_lock_file(os.path.join(lock_dir, f"session-{pid}.json"), ptr) + return ils.read_lock_file(path) or lock + + +def _apply_kwargs(repo, lock, lock_dir, **overrides): + kwargs = { + "remote": REMOTE, + "org": ORG, + "repo": REPO, + "issue_number": ISSUE, + "branch_name": BRANCH, + "worktree_path": repo["worktree"], + "claimant_identity": IDENTITY, + "claimant_profile": PROFILE, + "old_pid": lock.get("session_pid") or lock.get("pid"), + "expected_local_head": repo["local_head"], + "expected_remote_head": repo["remote_head"], + "expected_dirty_paths": repo["dirty_paths"], + "expected_fingerprints": repo["fingerprints"], + "existing_lock": lock, + "current_identity": IDENTITY, + "current_profile": PROFILE, + "role_kind": "author", + "current_pid": os.getpid(), + "current_branch": BRANCH, + "local_head": repo["local_head"], + "remote_head": repo["remote_head"], + "dirty_inventory": repo["inventory"], + "competing_live_locks": [], + "competing_sessions": [], + "workflow_lease_active": False, + "repo_root": repo["root"], + "dry_run": False, + "lock_dir": lock_dir, + } + kwargs.update(overrides) + return kwargs + + +# ── 1. Successful dead-PID same-claimant dirty rebind ─────────────────────── + + +def test_successful_dead_pid_same_claimant_dirty_rebind(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert result["success"], result + assert result["outcome"] == rebind.REBIND_SANCTIONED + assert result["old_pid"] == old + assert result["new_pid"] == os.getpid() + assert result["generation_after"] == result["generation_before"] + 1 + + rebound = ils.read_lock_file(result["lock_path"]) + assert rebound is not None + assert int(rebound["session_pid"]) == os.getpid() + assert int(rebound["pid"]) == os.getpid() + assert ( + rebound.get("lock_provenance", {}).get("source") + == issue_lock_provenance.SOURCE_DIRTY_SAME_CLAIMANT_REBIND + ) + assert rebound.get("rebind_record", {}).get("old_pid") == old + + +# ── 2. Byte-for-byte preservation ─────────────────────────────────────────── + + +def test_byte_for_byte_preservation(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + before = { + rel: rebind.content_fingerprint(os.path.join(dirty_repo["worktree"], rel)) + for rel in dirty_repo["dirty_paths"] + } + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert result["success"], result + after = { + rel: rebind.content_fingerprint(os.path.join(dirty_repo["worktree"], rel)) + for rel in dirty_repo["dirty_paths"] + } + assert before == after + assert result["fingerprints"] == before + + +# ── 3. Exact dirty-path and fingerprint enforcement ───────────────────────── + + +def test_extra_dirty_path_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + pins = list(dirty_repo["dirty_paths"])[:-1] # missing one observed path + fps = {p: dirty_repo["fingerprints"][p] for p in pins} + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + expected_dirty_paths=pins, + expected_fingerprints=fps, + ) + ) + assert not result["success"] + assert any("unexpected paths" in r for r in result["reasons"]) + + +def test_missing_expected_dirty_path_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + pins = list(dirty_repo["dirty_paths"]) + ["not_really_dirty.txt"] + fps = dict(dirty_repo["fingerprints"]) + fps["not_really_dirty.txt"] = "0" * 64 + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + expected_dirty_paths=pins, + expected_fingerprints=fps, + ) + ) + assert not result["success"] + assert any("missing expected" in r for r in result["reasons"]) + + +def test_modified_fingerprint_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + fps = dict(dirty_repo["fingerprints"]) + victim = dirty_repo["dirty_paths"][0] + fps[victim] = "f" * 64 + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + expected_fingerprints=fps, + ) + ) + assert not result["success"] + assert any("fingerprint disagreement" in r for r in result["reasons"]) + + +# ── 4. Atomic session-pointer replacement ─────────────────────────────────── + + +def test_session_pointer_points_to_lock(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert result["success"], result + new_ptr_path = os.path.join(lock_dir, f"session-{os.getpid()}.json") + assert os.path.exists(new_ptr_path) + ptr = ils.read_lock_file(new_ptr_path) + assert ptr is not None + assert os.path.realpath(ptr["lock_file_path"]) == os.path.realpath( + result["lock_path"] + ) + # Old pointer removed when it targeted this lock. + old_ptr = os.path.join(lock_dir, f"session-{old}.json") + assert not os.path.exists(old_ptr) + assert result.get("removed_old_session_pointer") is True + + +# ── 5. Retry after interruption (journal mid-state) ───────────────────────── + + +def test_retry_after_journal_mid_state(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + # Complete operation identity required for mid-flight resume (#868 F2). + rebind._atomic_write_json( + jpath, + { + "phase": rebind.JOURNAL_PHASE_PRE_BIND, + "issue_number": ISSUE, + "old_pid": old, + "new_pid": os.getpid(), + "expected_generation": 1, + "remote": REMOTE, + "org": ORG, + "repo": REPO, + "claimant_identity": IDENTITY, + "claimant_profile": PROFILE, + "source": rebind.SOURCE, + }, + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert result["success"], result + assert result["journal_phase"] == rebind.JOURNAL_PHASE_COMPLETE + + # Second apply is already_rebound (retry-safe). + rebound_lock = ils.read_lock_file(result["lock_path"]) + result2 = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + rebound_lock, + lock_dir, + old_pid=old, + existing_lock=rebound_lock, + ) + ) + assert result2["success"], result2 + assert result2["already_rebound"] is True + + +# ── 6. Active-PID refusal ─────────────────────────────────────────────────── + + +def test_active_pid_refused(dirty_repo, lock_dir): + live = os.getpid() + # Use a different "current" identity of session via fake current_pid... + # Owner is live (this process). Rebind must refuse. + lock = _make_lock(worktree=dirty_repo["worktree"], pid=live, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=live, + current_pid=live + 10_000_000, # distinct "new" session id for pin check + ) + ) + assert not result["success"] + assert any("still alive" in r for r in result["reasons"]) + + +# ── 7. Foreign claimant refusal ───────────────────────────────────────────── + + +def test_foreign_claimant_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + current_identity="someone-else", + claimant_identity="someone-else", + ) + ) + assert not result["success"] + assert any("foreign claimant" in r or "does not match" in r for r in result["reasons"]) + + +# ── 8. Profile mismatch refusal ───────────────────────────────────────────── + + +def test_profile_mismatch_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + current_profile="other-profile", + claimant_profile="other-profile", + ) + ) + assert not result["success"] + assert any("profile" in r for r in result["reasons"]) + + +# ── 9. Competing session/lock/lease refusal ───────────────────────────────── + + +def test_competing_live_lock_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + competing = [ + { + "issue_number": ISSUE, + "branch_name": BRANCH, + "worktree_path": dirty_repo["worktree"] + "-other", + "pid": os.getpid(), + } + ] + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + competing_live_locks=competing, + ) + ) + assert not result["success"] + assert any("competing live lock" in r for r in result["reasons"]) + + +def test_competing_session_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + competing_sessions=[ + {"pid": os.getpid(), "lock_file_path": lock["lock_file_path"], "live": True} + ], + ) + ) + # current_pid is os.getpid(), so same session is skipped — use another live pid. + # Spawn a long-lived process to act as competing live session. + rival = subprocess.Popen([sys.executable, "-c", "import time; time.sleep(30)"]) + try: + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + competing_sessions=[ + { + "pid": rival.pid, + "lock_file_path": lock["lock_file_path"], + "live": True, + } + ], + ) + ) + assert not result["success"] + assert any("competing live session" in r for r in result["reasons"]) + finally: + rival.kill() + rival.wait() + + +def test_workflow_lease_active_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + workflow_lease_active=True, + ) + ) + assert not result["success"] + assert any("workflow lease" in r for r in result["reasons"]) + + +# ── 10. Local- and remote-head movement refusal ───────────────────────────── + + +def test_local_head_movement_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + expected_local_head="b" * 40, + ) + ) + assert not result["success"] + assert any("local head" in r for r in result["reasons"]) + + +def test_remote_head_movement_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + expected_remote_head="c" * 40, + ) + ) + assert not result["success"] + assert any("remote head" in r for r in result["reasons"]) + + +# ── 11. Path/symlink/registration mismatches ──────────────────────────────── + + +def test_worktree_not_under_branches_refused(dirty_repo, lock_dir, tmp_path): + old = dead_pid() + # Use a path outside branches/ as the declared worktree (still real dir). + outside = tmp_path / "outside-wt" + outside.mkdir() + lock = _make_lock(worktree=str(outside), pid=old, lock_dir=lock_dir) + # Inventory empty for outside path; use empty pins to hit path gate first + # by providing matching empty-ish inventory after we force path checks. + inv = { + "dirty_paths": dirty_repo["dirty_paths"], + "fingerprints": dirty_repo["fingerprints"], + "ok": True, + "reasons": [], + } + result = rebind.assess_dirty_same_claimant_session_rebind( + remote=REMOTE, + org=ORG, + repo=REPO, + issue_number=ISSUE, + branch_name=BRANCH, + worktree_path=str(outside), + claimant_identity=IDENTITY, + claimant_profile=PROFILE, + old_pid=old, + expected_local_head=dirty_repo["local_head"], + expected_remote_head=dirty_repo["remote_head"], + expected_dirty_paths=dirty_repo["dirty_paths"], + expected_fingerprints=dirty_repo["fingerprints"], + existing_lock=lock, + current_identity=IDENTITY, + current_profile=PROFILE, + role_kind="author", + current_pid=os.getpid(), + current_branch=BRANCH, + local_head=dirty_repo["local_head"], + remote_head=dirty_repo["remote_head"], + dirty_inventory=inv, + repo_root=dirty_repo["root"], + ) + assert not result["rebind_sanctioned"] + assert any("branches/" in r for r in result["reasons"]) + + +def test_lock_worktree_mismatch_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock( + worktree=dirty_repo["worktree"] + "-elsewhere", + pid=old, + lock_dir=lock_dir, + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("does not match declared" in r for r in result["reasons"]) + + +# ── 12. Malformed lock/session records ────────────────────────────────────── + + +def test_malformed_lock_missing_pid_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + lock.pop("session_pid", None) + lock.pop("pid", None) + ils.save_lock_file(lock["lock_file_path"], lock) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("incomplete" in r or "session_pid" in r for r in result["reasons"]) + + +def test_empty_old_pid_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=None) + ) + assert not result["success"] + assert any("old_pid" in r for r in result["reasons"]) + + +def test_reviewer_role_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old, role_kind="reviewer") + ) + assert not result["success"] + assert any("reviewer" in r for r in result["reasons"]) + + +def test_reconciler_without_authorize_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + role_kind="reconciler", + authorize_reconciler_execute=False, + ) + ) + assert not result["success"] + assert any("authorize_reconciler_execute" in r for r in result["reasons"]) + + +# ── 13. No duplicate ownership after success or retry ─────────────────────── + + +def test_no_duplicate_ownership_after_success_or_retry(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + r1 = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert r1["success"], r1 + rebound = ils.read_lock_file(r1["lock_path"]) + r2 = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, rebound, lock_dir, old_pid=old, existing_lock=rebound) + ) + assert r2["success"], r2 + assert r2["already_rebound"] is True + + # Only one durable lock file for this issue; session pointer is current pid. + # Skip session pointers and rebind journals (dotfiles / non-lock records). + matching = [] + for p in ils.iter_lock_files(lock_dir): + name = os.path.basename(p) + if name.startswith(".") or name.startswith("session-"): + continue + rec = ils.read_lock_file(p) + if not rec: + continue + if ( + rec.get("issue_number") == ISSUE + and rec.get("remote") == REMOTE + and rec.get("branch_name") == BRANCH + and rec.get("session_pid") is not None + ): + matching.append(rec) + assert len(matching) == 1 + assert int(matching[0]["session_pid"]) == os.getpid() + # No live session pointer for the dead old pid. + assert not os.path.exists(os.path.join(lock_dir, f"session-{old}.json")) + + +# ── 14. Ordinary dirty-worktree locking remains fail-closed ───────────────── + + +def test_ordinary_dirty_lock_worktree_assessment_blocks(dirty_repo): + porcelain = dirty_repo["inventory"]["porcelain_status"] + assessment = issue_lock_worktree.assess_issue_lock_worktree( + worktree_path=dirty_repo["worktree"], + current_branch=BRANCH, + porcelain_status=porcelain, + base_equivalent=False, + ) + assert assessment["block"] is True + assert any( + "tracked file edits exist before issue lock" in r + for r in assessment["reasons"] + ) + + +# ── 15. Fixture matching #860 class with 7 fingerprint-pinned dirty paths ─── + + +def test_issue_860_regression_fixture_spec(): + spec = rebind.build_issue_860_regression_fixture_spec() + assert spec["claimant_identity"] == "jcwalker3" + assert spec["claimant_profile"] == "prgs-author" + assert spec["old_pid_alive"] is False + assert spec["live_session_pointer"] is None + assert spec["dirty_path_count"] == 7 + assert len(spec["expected_dirty_paths"]) == 7 + assert len(spec["expected_fingerprints"]) == 7 + assert spec["expected_local_head"] == spec["expected_remote_head"] + for path in spec["expected_dirty_paths"]: + assert path in spec["expected_fingerprints"] + assert len(spec["expected_fingerprints"][path]) == 64 + + +def test_dry_run_does_not_write(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + before = ils.read_lock_file(lock["lock_file_path"]) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old, dry_run=True) + ) + assert result["success"], result + assert result["dry_run"] is True + after = ils.read_lock_file(lock["lock_file_path"]) + assert after["session_pid"] == before["session_pid"] + assert not os.path.exists(os.path.join(lock_dir, f"session-{os.getpid()}.json")) + + +def test_provenance_source_is_sanctioned(): + assert ( + issue_lock_provenance.SOURCE_DIRTY_SAME_CLAIMANT_REBIND + in issue_lock_provenance.SANCTIONED_LOCK_SOURCES + ) + assessment = issue_lock_provenance.assess_lock_file_for_create_pr( + { + "work_lease": {"operation_type": "author_issue_work"}, + "lock_provenance": { + "source": issue_lock_provenance.SOURCE_DIRTY_SAME_CLAIMANT_REBIND, + "written_by_tool": issue_lock_provenance.SOURCE_DIRTY_SAME_CLAIMANT_REBIND, + "written_at": "2026-01-01T00:00:00Z", + }, + } + ) + assert assessment["proven"] is True + + +def test_permission_allowed_is_not_ownership_proof(dirty_repo, lock_dir): + """permission_allowed=True must not bypass foreign claimant refusal.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.assess_dirty_same_claimant_session_rebind( + remote=REMOTE, + org=ORG, + repo=REPO, + issue_number=ISSUE, + branch_name=BRANCH, + worktree_path=dirty_repo["worktree"], + claimant_identity=IDENTITY, + claimant_profile=PROFILE, + old_pid=old, + expected_local_head=dirty_repo["local_head"], + expected_remote_head=dirty_repo["remote_head"], + expected_dirty_paths=dirty_repo["dirty_paths"], + expected_fingerprints=dirty_repo["fingerprints"], + existing_lock=lock, + current_identity="intruder", + current_profile=PROFILE, + role_kind="author", + current_pid=os.getpid(), + current_branch=BRANCH, + local_head=dirty_repo["local_head"], + remote_head=dirty_repo["remote_head"], + dirty_inventory=dirty_repo["inventory"], + permission_allowed=True, + repo_root=dirty_repo["root"], + ) + assert not result["rebind_sanctioned"] + assert any("does not match active identity" in r for r in result["reasons"]) + + +def test_content_fingerprint_stable(tmp_path): + p = tmp_path / "f.txt" + p.write_bytes(b"abc123") + a = rebind.content_fingerprint(str(p)) + b = rebind.content_fingerprint(str(p)) + assert a == b + assert len(a) == 64 + + +# ── #868 F1 — Complete dirty-inventory revalidation ───────────────────────── + + +def test_extra_tracked_dirty_path_before_bind_refused(dirty_repo, lock_dir): + """Extra tracked dirty path appearing immediately before binding fails closed.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + wt = dirty_repo["worktree"] + # Seed a tracked file, then dirty it without including it in the pin set. + tracked_extra = "tracked_extra_before_bind.txt" + path = Path(wt) / tracked_extra + path.write_text("seed tracked extra\n", encoding="utf-8") + _git(wt, "add", tracked_extra) + _git(wt, "commit", "-q", "-m", "seed extra tracked") + # Heads moved — re-pin heads so only inventory disagreement is tested. + head = _git(wt, "rev-parse", "HEAD").stdout.strip() + _git(wt, "push", "-q", "origin", BRANCH) + remote_head = _git(wt, "rev-parse", f"refs/remotes/origin/{BRANCH}").stdout.strip() + path.write_text("dirty tracked extra\n", encoding="utf-8") + # Pins still describe the original inventory (without tracked_extra). + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + expected_local_head=head, + expected_remote_head=remote_head, + local_head=head, + remote_head=remote_head, + dirty_inventory=None, # force live recollect in apply + ) + ) + assert not result["success"] + joined = " ".join(result["reasons"]) + assert "unexpected paths" in joined or "path-set disagreement" in joined + # Must not leave a newly authoritative live session for this pid. + rebound = ils.read_lock_file(lock["lock_file_path"]) + assert rebound is not None + assert int(rebound.get("session_pid") or 0) == old + + +def test_extra_untracked_path_before_bind_refused(dirty_repo, lock_dir): + """Extra untracked path appearing immediately before binding fails closed.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + wt = dirty_repo["worktree"] + extra = Path(wt) / "surprise_untracked_before_bind.txt" + extra.write_text("sneaky\n", encoding="utf-8") + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + dirty_inventory=None, + ) + ) + assert not result["success"] + joined = " ".join(result["reasons"]) + assert "unexpected paths" in joined or "path-set disagreement" in joined + rebound = ils.read_lock_file(lock["lock_file_path"]) + assert int(rebound.get("session_pid") or 0) == old + + +def test_path_added_during_mutation_window_refused(dirty_repo, lock_dir, monkeypatch): + """Path added during the mutation window is detected by post-bind inventory.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + wt = dirty_repo["worktree"] + real_bind = ils.bind_session_lock + + def _bind_then_add_path(lock_payload, **kwargs): + path = real_bind(lock_payload, **kwargs) + surprise = Path(wt) / "added_during_bind.txt" + surprise.write_text("during bind\n", encoding="utf-8") + return path + + monkeypatch.setattr(rebind, "bind_session_lock", _bind_then_add_path) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old, dirty_inventory=None) + ) + assert not result["success"] + joined = " ".join(result["reasons"]) + assert "post-bind" in joined + assert "unexpected paths" in joined or "path-set disagreement" in joined + assert result.get("journal_phase") == "post_bind_inventory_failed" + + +def test_path_removed_during_mutation_window_refused(dirty_repo, lock_dir, monkeypatch): + """Path removed during the mutation window is detected by post-bind inventory.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + wt = dirty_repo["worktree"] + victim = dirty_repo["dirty_paths"][-1] # prefer untracked for easy remove + real_bind = ils.bind_session_lock + + def _bind_then_remove_path(lock_payload, **kwargs): + path = real_bind(lock_payload, **kwargs) + target = Path(wt) / victim + if target.exists(): + target.unlink() + return path + + monkeypatch.setattr(rebind, "bind_session_lock", _bind_then_remove_path) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old, dirty_inventory=None) + ) + assert not result["success"] + joined = " ".join(result["reasons"]) + assert "post-bind" in joined + assert "missing expected" in joined or "path-set disagreement" in joined + + +def test_fingerprint_movement_unchanged_path_set_refused(dirty_repo, lock_dir, monkeypatch): + """Fingerprint movement with unchanged path set fails pre- or post-bind check.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + wt = dirty_repo["worktree"] + victim = dirty_repo["dirty_paths"][0] + real_bind = ils.bind_session_lock + + def _bind_then_mutate_bytes(lock_payload, **kwargs): + path = real_bind(lock_payload, **kwargs) + target = Path(wt) / victim + target.write_text(target.read_text(encoding="utf-8") + "mutated\n", encoding="utf-8") + return path + + monkeypatch.setattr(rebind, "bind_session_lock", _bind_then_mutate_bytes) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old, dirty_inventory=None) + ) + assert not result["success"] + joined = " ".join(result["reasons"]) + assert "fingerprint" in joined + assert "post-bind" in joined + + +# ── #868 F2 — Complete recovery-journal identity ──────────────────────────── + + +def _complete_journal(**overrides): + base = { + "phase": rebind.JOURNAL_PHASE_ASSESSED, + "issue_number": ISSUE, + "branch_name": BRANCH, + "worktree_path": "/tmp/wt", + "old_pid": 1, + "new_pid": os.getpid(), + "remote": REMOTE, + "org": ORG, + "repo": REPO, + "claimant_identity": IDENTITY, + "claimant_profile": PROFILE, + "source": rebind.SOURCE, + } + base.update(overrides) + return base + + +def test_journal_remote_mismatch_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + rebind._atomic_write_json( + jpath, + _complete_journal( + phase=rebind.JOURNAL_PHASE_PRE_BIND, + old_pid=old, + remote="dadeschools", + ), + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("remote" in r and "mismatch" in r for r in result["reasons"]) + + +def test_journal_org_mismatch_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + rebind._atomic_write_json( + jpath, + _complete_journal( + phase=rebind.JOURNAL_PHASE_PRE_BIND, + old_pid=old, + org="Other-Org", + ), + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("org" in r and "mismatch" in r for r in result["reasons"]) + + +def test_journal_repo_mismatch_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + rebind._atomic_write_json( + jpath, + _complete_journal( + phase=rebind.JOURNAL_PHASE_PRE_BIND, + old_pid=old, + repo="Other-Repo", + ), + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("repo" in r and "mismatch" in r for r in result["reasons"]) + + +def test_journal_claimant_identity_mismatch_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + rebind._atomic_write_json( + jpath, + _complete_journal( + phase=rebind.JOURNAL_PHASE_PRE_BIND, + old_pid=old, + claimant_identity="intruder", + ), + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("claimant_identity" in r for r in result["reasons"]) + + +def test_journal_claimant_profile_mismatch_refused(dirty_repo, lock_dir): + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + rebind._atomic_write_json( + jpath, + _complete_journal( + phase=rebind.JOURNAL_PHASE_PRE_BIND, + old_pid=old, + claimant_profile="prgs-reviewer", + ), + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("claimant_profile" in r for r in result["reasons"]) + + +def test_incomplete_legacy_journal_identity_refused(dirty_repo, lock_dir): + """Pre-#868 journals missing the five identity fields fail closed mid-flight.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + rebind._atomic_write_json( + jpath, + { + "phase": rebind.JOURNAL_PHASE_PRE_BIND, + "issue_number": ISSUE, + "old_pid": old, + "new_pid": os.getpid(), + # deliberately omit remote/org/repo/claimant_* + }, + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("omits operation identity" in r for r in result["reasons"]) + + +def test_journal_replay_cross_repository_refused(dirty_repo, lock_dir): + """Replaying a journal from another repository is refused.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + rebind._atomic_write_json( + jpath, + _complete_journal( + phase=rebind.JOURNAL_PHASE_ASSESSED, + old_pid=old, + remote="dadeschools", + org="Other-Org", + repo="Other-Repo", + ), + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("replay" in r or "mismatch" in r for r in result["reasons"]) + + +def test_journal_replay_cross_claimant_refused(dirty_repo, lock_dir): + """Replaying a journal from another claimant is refused.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + jpath = rebind.journal_path(lock_dir, ISSUE) + rebind._atomic_write_json( + jpath, + _complete_journal( + phase=rebind.JOURNAL_PHASE_ASSESSED, + old_pid=old, + claimant_identity="other-user", + claimant_profile="other-profile", + ), + ) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert not result["success"] + assert any("claimant" in r for r in result["reasons"]) + + +def test_successful_exact_retry_with_complete_identity(dirty_repo, lock_dir): + """Exact retry after success is already_rebound with complete matching identity.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + r1 = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert r1["success"], r1 + # Journal must persist the five identity fields. + jpath = rebind.journal_path(lock_dir, ISSUE) + journal = rebind._read_json(jpath) + assert journal is not None + assert journal["phase"] == rebind.JOURNAL_PHASE_COMPLETE + for field in rebind.REQUIRED_JOURNAL_IDENTITY_FIELDS: + assert journal.get(field), field + assert journal["remote"] == REMOTE + assert journal["org"] == ORG + assert journal["repo"] == REPO + assert journal["claimant_identity"] == IDENTITY + assert journal["claimant_profile"] == PROFILE + + rebound = ils.read_lock_file(r1["lock_path"]) + r2 = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + rebound, + lock_dir, + old_pid=old, + existing_lock=rebound, + ) + ) + assert r2["success"], r2 + assert r2["already_rebound"] is True + + +def test_already_rebound_requires_complete_matching_identity(dirty_repo, lock_dir): + """already_rebound with mismatched journal identity fails closed.""" + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + r1 = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs(dirty_repo, lock, lock_dir, old_pid=old) + ) + assert r1["success"], r1 + # Corrupt journal identity after success. + jpath = rebind.journal_path(lock_dir, ISSUE) + journal = rebind._read_json(jpath) + assert journal is not None + journal["claimant_identity"] = "not-the-owner" + rebind._atomic_write_json(jpath, journal) + + rebound = ils.read_lock_file(r1["lock_path"]) + r2 = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + rebound, + lock_dir, + old_pid=old, + existing_lock=rebound, + ) + ) + # Same pid on lock would look like already_rebound, but identity must match. + assert not r2["success"] + assert any("claimant_identity" in r or "mismatch" in r for r in r2["reasons"]) + + +def test_ordinary_dirty_worktree_refusal_preserved(dirty_repo): + """#868 must not weaken ordinary dirty-worktree refusal on lock_issue path.""" + porcelain = dirty_repo["inventory"]["porcelain_status"] + assessment = issue_lock_worktree.assess_issue_lock_worktree( + worktree_path=dirty_repo["worktree"], + current_branch=BRANCH, + porcelain_status=porcelain, + base_equivalent=False, + ) + assert assessment["block"] is True + + +# ── #868 — Reconciler success path ────────────────────────────────────────── + + +def test_reconciler_success_path_tightly_pinned(dirty_repo, lock_dir): + """Reconciler with authorize_reconciler_execute=True may execute rebind. + + Reconciler execution grants no commit/push/publication/review/merge + capability — only the tightly pinned session rebind. + """ + old = dead_pid() + lock = _make_lock(worktree=dirty_repo["worktree"], pid=old, lock_dir=lock_dir) + result = rebind.apply_dirty_same_claimant_session_rebind( + **_apply_kwargs( + dirty_repo, + lock, + lock_dir, + old_pid=old, + role_kind="reconciler", + authorize_reconciler_execute=True, + # Reconciler may act for the recorded claimant without being that + # identity in the active session (still pin-checked against lock). + current_identity="sysadmin", + current_profile="prgs-reconciler", + ) + ) + assert result["success"], result + assert result["outcome"] == rebind.REBIND_SANCTIONED + rebound = ils.read_lock_file(result["lock_path"]) + assert int(rebound["session_pid"]) == os.getpid() + # Provenance records the rebind tool; no publication authority is granted. + assert ( + rebound.get("lock_provenance", {}).get("source") + == issue_lock_provenance.SOURCE_DIRTY_SAME_CLAIMANT_REBIND + ) + # Reconciler rebind does not stamp commit/push/review/merge capabilities. + prov = rebound.get("lock_provenance") or {} + blob = json.dumps(prov) + for forbidden in ( + "gitea.repo.commit", + "gitea.branch.push", + "gitea.pr.approve", + "gitea.pr.merge", + "gitea.pr.create", + ): + assert forbidden not in blob + + +def test_revalidate_complete_dirty_inventory_helper(dirty_repo): + ok = rebind.revalidate_complete_dirty_inventory( + dirty_repo["worktree"], + expected_dirty_paths=dirty_repo["dirty_paths"], + expected_fingerprints=dirty_repo["fingerprints"], + phase="unit", + ) + assert ok["ok"] is True + + bad = rebind.revalidate_complete_dirty_inventory( + dirty_repo["worktree"], + expected_dirty_paths=dirty_repo["dirty_paths"][:-1], + expected_fingerprints={ + p: dirty_repo["fingerprints"][p] for p in dirty_repo["dirty_paths"][:-1] + }, + phase="unit", + ) + assert bad["ok"] is False + assert any("unexpected paths" in r for r in bad["reasons"]) + + +def test_validate_journal_operation_identity_helper(): + complete = _complete_journal() + assert ( + rebind.validate_journal_operation_identity( + complete, + remote=REMOTE, + org=ORG, + repo=REPO, + claimant_identity=IDENTITY, + claimant_profile=PROFILE, + ) + == [] + ) + incomplete = {"phase": "assessed", "remote": REMOTE} + reasons = rebind.validate_journal_operation_identity( + incomplete, + remote=REMOTE, + org=ORG, + repo=REPO, + claimant_identity=IDENTITY, + claimant_profile=PROFILE, + ) + assert any("omits operation identity" in r for r in reasons) + assert any("org" in r for r in reasons) diff --git a/tests/test_issue_855_expired_reviewer_reclaim.py b/tests/test_issue_855_expired_reviewer_reclaim.py new file mode 100644 index 0000000..288b7f7 --- /dev/null +++ b/tests/test_issue_855_expired_reviewer_reclaim.py @@ -0,0 +1,244 @@ +"""#855 AC4: an expired reviewer lease must not indefinitely protect an +already-merged branch when no live claimant exists. + +Two layers are covered: + +* ``branch_cleanup_guard.assess_expired_reviewer_lease_reclaim`` — the pure, + fail-closed reclaim decision. Every condition must be provably satisfied or + the lease keeps protecting the branch. +* ``gitea_mcp_server._collect_branch_ownership_records`` — the wiring that + supplies authoritative evidence (PR merged state, owner-process liveness, + competing ownership) to that decision, and flips an expired reviewer lease + to reclaimable only under the full policy. + +All inputs are fabricated; no real repository, lease, or credential is used. +""" + +import importlib +import unittest +from unittest.mock import patch + +import branch_cleanup_guard + +mcp_server = importlib.import_module("gitea_mcp_server") + +FAKE_AUTH = "token fake" +REMOTE = "prgs" +ORG = "Scaled-Tech-Consulting" +REPO = "Gitea-Tools" +HOST = "gitea.prgs.cc" +BRANCH = "feat/issue-638-webui-app-shell-phase1" +PR_NUMBER = 818 + + +class TestAssessExpiredReviewerLeaseReclaim(unittest.TestCase): + """Pure fail-closed reclaim decision (#855 AC4).""" + + def _call(self, **overrides): + base = dict( + role="reviewer", + status="expired", + pr_merged=True, + owner_pid_alive=False, + competing_active_claimant=False, + ) + base.update(overrides) + return branch_cleanup_guard.assess_expired_reviewer_lease_reclaim(**base) + + def test_full_policy_satisfied_allows_reclaim(self): + out = self._call() + self.assertTrue(out["reclaim_allowed"]) + self.assertEqual(out["reasons"], []) + self.assertEqual(out["decision"], "reclaim_expired_reviewer_lease") + + def test_stale_dead_process_reviewer_also_reclaimable(self): + out = self._call(status="stale_dead_process") + self.assertTrue(out["reclaim_allowed"]) + + def test_non_reviewer_role_never_reclaims(self): + for role in ("author", "merger", "controller", "reconciler", "unknown"): + with self.subTest(role=role): + out = self._call(role=role) + self.assertFalse(out["reclaim_allowed"]) + self.assertTrue(out["reasons"]) + self.assertEqual(out["decision"], "keep_protecting") + + def test_active_status_never_reclaims(self): + out = self._call(status="active") + self.assertFalse(out["reclaim_allowed"]) + + def test_pr_not_merged_blocks_reclaim(self): + out = self._call(pr_merged=False) + self.assertFalse(out["reclaim_allowed"]) + + def test_pr_merged_unknown_fails_closed(self): + out = self._call(pr_merged=None) + self.assertFalse(out["reclaim_allowed"]) + + def test_owner_process_alive_blocks_reclaim(self): + out = self._call(owner_pid_alive=True) + self.assertFalse(out["reclaim_allowed"]) + + def test_owner_liveness_unknown_fails_closed(self): + out = self._call(owner_pid_alive=None) + self.assertFalse(out["reclaim_allowed"]) + + def test_competing_active_claimant_blocks_reclaim(self): + out = self._call(competing_active_claimant=True) + self.assertFalse(out["reclaim_allowed"]) + + def test_competing_claimant_unknown_fails_closed(self): + out = self._call(competing_active_claimant=None) + self.assertFalse(out["reclaim_allowed"]) + + def test_reasons_never_leak_secrets(self): + out = self._call(role="author") + blob = " ".join(out["reasons"]).lower() + self.assertNotIn("token", blob) + self.assertNotIn("password", blob) + + +class _FakeLease(dict): + pass + + +class TestCollectorExpiredReviewerReclaimWiring(unittest.TestCase): + """`_collect_branch_ownership_records` supplies authoritative evidence and + flips an expired reviewer lease to reclaimable only under the full policy.""" + + def _run( + self, + *, + lease_role="reviewer", + lease_freshness="stale_dead_process", + owner_pid_alive=False, + pr_merged=True, + extra_leases=None, + worktree_on_branch=False, + ): + lease = _FakeLease( + role=lease_role, + work_kind="pr", + work_number=PR_NUMBER, + branch=BRANCH, + status="active", + owner_pid=999999, + remote=REMOTE, + org=ORG, + repo=REPO, + host=HOST, + freshness={ + "freshness": lease_freshness, + "owner_pid": 999999, + "owner_pid_alive": owner_pid_alive, + "expired_by_time": lease_freshness == "expired", + }, + ) + leases = [lease] + list(extra_leases or []) + + pr_payload = { + "number": PR_NUMBER, + "merged": pr_merged, + "merged_at": "2026-07-23T00:00:00Z" if pr_merged else None, + "head": {"ref": BRANCH}, + } + + def fake_api_request(method, url, *a, **k): + if method == "GET" and f"/pulls/{PR_NUMBER}" in url: + return pr_payload + raise AssertionError(f"unexpected api_request {method} {url}") + + wt_entries = [] + if worktree_on_branch: + wt_entries = [{"branch": BRANCH, "path": f"/x/branches/{BRANCH}"}] + + with patch.object( + mcp_server.lease_lifecycle, + "list_active_leases", + return_value={"leases": leases}, + ), patch.object( + mcp_server.control_plane_db, "get_db", return_value=object(), create=True + ), patch.object( + mcp_server.issue_lock_store, "iter_lock_files", return_value=[] + ), patch.object( + mcp_server.worktree_cleanup_audit, + "list_worktrees", + return_value=wt_entries, + ), patch.object( + mcp_server, "api_get_all", return_value=[] + ), patch.object( + mcp_server, "api_request", side_effect=fake_api_request + ): + return mcp_server._collect_branch_ownership_records( + remote=REMOTE, + host=HOST, + org=ORG, + repo=REPO, + branch=BRANCH, + pr_number=PR_NUMBER, + project_root="/x", + auth=FAKE_AUTH, + base_api="https://gitea.prgs.cc/api/v1/repos/x/y", + ) + + def _reviewer_records(self, bundle): + return [ + rec + for rec in bundle["records"] + if rec.get("category") + == branch_cleanup_guard.OWNERSHIP_CATEGORY_REVIEWER_LEASE + ] + + def test_merged_dead_uncontested_reviewer_lease_is_reclaimable(self): + bundle = self._run() + self.assertFalse(bundle["inventory_error"]) + recs = self._reviewer_records(bundle) + self.assertEqual(len(recs), 1) + self.assertTrue(recs[0]["reclaim_allowed"]) + # And the guard consequently does not block deletion on it. + ownership = branch_cleanup_guard.assess_active_branch_ownership( + remote=REMOTE, org=ORG, repo=REPO, branch=BRANCH, host=HOST, + records=bundle["records"], + ) + self.assertFalse(ownership["block"]) + + def test_unmerged_pr_keeps_reviewer_lease_protective(self): + bundle = self._run(pr_merged=False) + recs = self._reviewer_records(bundle) + self.assertEqual(len(recs), 1) + self.assertFalse(recs[0]["reclaim_allowed"]) + ownership = branch_cleanup_guard.assess_active_branch_ownership( + remote=REMOTE, org=ORG, repo=REPO, branch=BRANCH, host=HOST, + records=bundle["records"], + ) + self.assertTrue(ownership["block"]) + + def test_owner_process_alive_keeps_reviewer_lease_protective(self): + bundle = self._run(owner_pid_alive=True, lease_freshness="expired") + recs = self._reviewer_records(bundle) + self.assertFalse(recs[0]["reclaim_allowed"]) + + def test_competing_worktree_binding_keeps_reviewer_lease_protective(self): + bundle = self._run(worktree_on_branch=True) + recs = self._reviewer_records(bundle) + self.assertFalse(recs[0]["reclaim_allowed"]) + ownership = branch_cleanup_guard.assess_active_branch_ownership( + remote=REMOTE, org=ORG, repo=REPO, branch=BRANCH, host=HOST, + records=bundle["records"], + ) + self.assertTrue(ownership["block"]) + + def test_expired_author_lease_never_reclaimed_by_reviewer_policy(self): + bundle = self._run(lease_role="author") + author_recs = [ + rec + for rec in bundle["records"] + if rec.get("category") + == branch_cleanup_guard.OWNERSHIP_CATEGORY_AUTHOR_LEASE + ] + self.assertEqual(len(author_recs), 1) + self.assertFalse(author_recs[0]["reclaim_allowed"]) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/test_issue_858_audit_merged_pr_aware.py b/tests/test_issue_858_audit_merged_pr_aware.py new file mode 100644 index 0000000..3232318 --- /dev/null +++ b/tests/test_issue_858_audit_merged_pr_aware.py @@ -0,0 +1,551 @@ +"""Merged-PR awareness for the worktree cleanup audit (#858). + +Before #858 an ``issue_work`` worktree could never leave ``active_issue_work``: +the audit had no PR linkage at all (``pr_number`` was structurally ``None``) +and its only route to ``clean_stale_removable`` was a TTL derived from a +``last_used_at`` that nothing ever populated. A merged, clean, unprotected +worktree was therefore reported as active work forever, disagreeing with the +PR-scoped reconciler. + +These tests use fabricated temporary repositories and synthetic PR records +only. Nothing here removes a worktree or deletes a branch. +""" + +import os +import subprocess +import sys +import tempfile +import unittest +from unittest.mock import patch + +sys.path.insert(0, str(__import__("pathlib").Path(__file__).resolve().parent.parent)) + +import merged_cleanup_reconcile as mcr # noqa: E402 +import worktree_cleanup_audit as wca # noqa: E402 + + +MERGED_BRANCH = "feat/issue-777-timeline" +MERGED_PATH = "/repo/branches/issue-777-timeline" +HEAD_SHA = "a" * 40 + + +def _pr(number, branch, *, merged=True, sha=HEAD_SHA, state=None): + """Synthetic Gitea PR payload.""" + return { + "number": number, + "head": {"ref": branch, "sha": sha}, + "merged_at": "2026-07-24T01:00:00Z" if merged else None, + "state": state or ("closed" if merged else "open"), + } + + +def _porcelain(*entries): + out = [] + for path, branch, sha in entries: + out.append(f"worktree {path}") + out.append(f"HEAD {sha}") + if branch is None: + out.append("detached") + else: + out.append(f"branch refs/heads/{branch}") + out.append("") + return "\n".join(out) + + +class _AuditHarness(unittest.TestCase): + """Runs audit_branches_directory over a fabricated worktree listing.""" + + PORCELAIN = _porcelain( + ("/repo", "master", "f" * 40), + (MERGED_PATH, MERGED_BRANCH, HEAD_SHA), + ) + + def run_audit(self, *, dirty_paths=(), contained=True, **kwargs): + def fake_dirty(path): + if path in dirty_paths: + return {"exists": True, "dirty": True, "dirty_files": [" M x.py"]} + return {"exists": True, "dirty": False, "dirty_files": []} + + with patch.object( + wca, "list_worktrees", + return_value=wca.parse_worktree_porcelain(self.PORCELAIN), + ), patch.object( + wca, "read_worktree_dirty", side_effect=fake_dirty + ), patch.object( + wca, "git_worktree_list", return_value="(mocked)" + ), patch.object( + wca, "is_head_ancestor_of_ref", return_value=contained + ): + report = wca.audit_branches_directory("/repo", **kwargs) + return {wt["path"]: wt for wt in report["worktrees"]}, report + + def merged_audit(self, **kwargs): + kwargs.setdefault("pr_index", wca.build_pr_index([_pr(849, MERGED_BRANCH)])) + kwargs.setdefault("master_ref", "prgs/master") + return self.run_audit(**kwargs) + + +class TestMergedWorktreeBecomesRemovable(_AuditHarness): + def test_clean_merged_issue_worktree_is_linked_and_removable(self): + by_path, report = self.merged_audit() + entry = by_path[MERGED_PATH] + + self.assertEqual(entry["classification"], wca.CLASS_CLEAN_STALE_REMOVABLE) + self.assertTrue(entry["removable"]) + self.assertEqual(entry["merged_pr_linkage"]["status"], wca.LINKAGE_MERGED) + self.assertEqual(entry["merged_pr_cleanup"]["block_reasons"], []) + self.assertIn(MERGED_PATH, [c["path"] for c in report["removable_candidates"]]) + + def test_pr_number_populated_from_authoritative_linkage(self): + by_path, _ = self.merged_audit() + self.assertEqual(by_path[MERGED_PATH]["pr_number"], 849) + + def test_regression_without_pr_evidence_stays_active_issue_work(self): + """The pre-#858 behaviour, still correct when no PR state is supplied.""" + by_path, _ = self.run_audit() + entry = by_path[MERGED_PATH] + self.assertEqual(entry["classification"], wca.CLASS_ACTIVE_ISSUE_WORK) + self.assertFalse(entry["removable"]) + self.assertIsNone(entry["pr_number"]) + + +class TestProtectiveSignalsSurvive(_AuditHarness): + def test_open_pr_worktree_is_not_removable(self): + index = wca.build_pr_index([_pr(900, MERGED_BRANCH, merged=False)]) + by_path, _ = self.run_audit( + pr_index=index, + master_ref="prgs/master", + open_pr_branches={MERGED_BRANCH}, + ) + entry = by_path[MERGED_PATH] + self.assertEqual(entry["classification"], wca.CLASS_ACTIVE_OPEN_PR) + self.assertFalse(entry["removable"]) + # linkage still reports the owning PR, it just is not merge proof + self.assertEqual(entry["merged_pr_linkage"]["status"], wca.LINKAGE_OPEN) + self.assertEqual(entry["pr_number"], 900) + + def test_dirty_tracked_worktree_is_not_removable(self): + by_path, _ = self.merged_audit(dirty_paths=(MERGED_PATH,)) + entry = by_path[MERGED_PATH] + self.assertEqual(entry["classification"], wca.CLASS_DIRTY_LOCAL) + self.assertFalse(entry["removable"]) + self.assertIn( + "worktree has uncommitted changes", + entry["merged_pr_cleanup"]["block_reasons"], + ) + + def test_untracked_only_worktree_is_not_removable(self): + """``git status --porcelain`` reports untracked files as dirty too.""" + def untracked(path): + if path == MERGED_PATH: + return {"exists": True, "dirty": True, "dirty_files": ["?? scratch.txt"]} + return {"exists": True, "dirty": False, "dirty_files": []} + + with patch.object( + wca, "list_worktrees", + return_value=wca.parse_worktree_porcelain(self.PORCELAIN), + ), patch.object( + wca, "read_worktree_dirty", side_effect=untracked + ), patch.object( + wca, "git_worktree_list", return_value="(mocked)" + ), patch.object( + wca, "is_head_ancestor_of_ref", return_value=True + ): + report = wca.audit_branches_directory( + "/repo", + pr_index=wca.build_pr_index([_pr(849, MERGED_BRANCH)]), + master_ref="prgs/master", + ) + entry = {wt["path"]: wt for wt in report["worktrees"]}[MERGED_PATH] + self.assertEqual(entry["classification"], wca.CLASS_DIRTY_LOCAL) + self.assertFalse(entry["removable"]) + + def test_active_lease_by_issue_number_is_protective(self): + by_path, _ = self.merged_audit(leased_issue_numbers={777}) + entry = by_path[MERGED_PATH] + self.assertTrue(entry["has_active_lease"]) + self.assertEqual(entry["classification"], wca.CLASS_ACTIVE_ISSUE_WORK) + self.assertFalse(entry["removable"]) + + def test_active_lease_by_branch_is_protective(self): + by_path, _ = self.merged_audit(leased_branches={MERGED_BRANCH}) + entry = by_path[MERGED_PATH] + self.assertTrue(entry["has_active_lease"]) + self.assertFalse(entry["removable"]) + + def test_active_issue_lock_is_protective(self): + by_path, _ = self.merged_audit(active_issue_branches={MERGED_BRANCH}) + entry = by_path[MERGED_PATH] + self.assertTrue(entry["has_active_issue_lock"]) + self.assertEqual(entry["classification"], wca.CLASS_ACTIVE_ISSUE_WORK) + self.assertFalse(entry["removable"]) + + def test_live_session_worktree_is_protective(self): + by_path, _ = self.merged_audit(live_session_paths={MERGED_PATH}) + entry = by_path[MERGED_PATH] + self.assertTrue(entry["has_live_session"]) + self.assertEqual(entry["classification"], wca.CLASS_ACTIVE_ISSUE_WORK) + self.assertFalse(entry["removable"]) + + def test_head_not_contained_in_master_is_not_removable(self): + by_path, _ = self.merged_audit(contained=False) + entry = by_path[MERGED_PATH] + self.assertEqual(entry["classification"], wca.CLASS_ACTIVE_ISSUE_WORK) + self.assertFalse(entry["removable"]) + self.assertIn( + "worktree head is not contained in authoritative master " + "(unmerged commits remain)", + entry["merged_pr_cleanup"]["block_reasons"], + ) + + def test_unknown_containment_fails_closed(self): + by_path, _ = self.merged_audit(contained=None) + entry = by_path[MERGED_PATH] + self.assertFalse(entry["removable"]) + self.assertIn( + "containment of the worktree head in master is unknown", + entry["merged_pr_cleanup"]["block_reasons"], + ) + + def test_missing_master_ref_fails_closed(self): + by_path, _ = self.run_audit( + pr_index=wca.build_pr_index([_pr(849, MERGED_BRANCH)]) + ) + self.assertFalse(by_path[MERGED_PATH]["removable"]) + + def test_unmerged_owning_pr_is_not_removable(self): + index = wca.build_pr_index([_pr(901, MERGED_BRANCH, merged=False)]) + by_path, _ = self.run_audit(pr_index=index, master_ref="prgs/master") + entry = by_path[MERGED_PATH] + self.assertFalse(entry["removable"]) + self.assertIn( + "owning PR #901 is not merged", + entry["merged_pr_cleanup"]["block_reasons"], + ) + + def test_control_checkout_is_never_removable(self): + by_path, _ = self.merged_audit() + control = by_path["/repo"] + self.assertTrue(control["is_protected"]) + self.assertEqual(control["classification"], wca.CLASS_UNSAFE_UNKNOWN) + self.assertFalse(control["removable"]) + + def test_control_checkout_not_removable_even_if_linked_and_merged(self): + """A merged PR on the control checkout must not unlock removal.""" + porcelain = _porcelain(("/repo", MERGED_BRANCH, HEAD_SHA)) + with patch.object( + wca, "list_worktrees", return_value=wca.parse_worktree_porcelain(porcelain) + ), patch.object( + wca, "read_worktree_dirty", + return_value={"exists": True, "dirty": False, "dirty_files": []}, + ), patch.object( + wca, "git_worktree_list", return_value="(mocked)" + ), patch.object( + wca, "is_head_ancestor_of_ref", return_value=True + ): + report = wca.audit_branches_directory( + "/repo", + pr_index=wca.build_pr_index([_pr(849, MERGED_BRANCH)]), + master_ref="prgs/master", + ) + entry = report["worktrees"][0] + self.assertEqual(entry["classification"], wca.CLASS_UNSAFE_UNKNOWN) + self.assertFalse(entry["removable"]) + + +class TestAmbiguousLinkageFailsClosed(_AuditHarness): + def test_competing_prs_on_one_branch_fail_closed(self): + index = wca.build_pr_index( + [_pr(849, MERGED_BRANCH), _pr(860, MERGED_BRANCH)] + ) + by_path, _ = self.run_audit(pr_index=index, master_ref="prgs/master") + entry = by_path[MERGED_PATH] + self.assertEqual(entry["merged_pr_linkage"]["status"], wca.LINKAGE_AMBIGUOUS) + self.assertIsNone(entry["pr_number"]) + self.assertEqual(entry["classification"], wca.CLASS_ACTIVE_ISSUE_WORK) + self.assertFalse(entry["removable"]) + + def test_merged_plus_open_pr_on_one_branch_fails_closed(self): + index = wca.build_pr_index( + [_pr(849, MERGED_BRANCH), _pr(861, MERGED_BRANCH, merged=False)] + ) + by_path, _ = self.run_audit(pr_index=index, master_ref="prgs/master") + entry = by_path[MERGED_PATH] + self.assertEqual(entry["merged_pr_linkage"]["status"], wca.LINKAGE_AMBIGUOUS) + self.assertFalse(entry["removable"]) + + def test_no_owning_pr_fails_closed(self): + by_path, _ = self.run_audit( + pr_index=wca.build_pr_index([_pr(849, "feat/other-branch")]), + master_ref="prgs/master", + ) + entry = by_path[MERGED_PATH] + self.assertEqual(entry["merged_pr_linkage"]["status"], wca.LINKAGE_NONE) + self.assertFalse(entry["removable"]) + + def test_malformed_pr_records_are_dropped_not_guessed(self): + index = wca.build_pr_index( + [ + {"number": None, "head": {"ref": MERGED_BRANCH}}, + {"number": 5, "head": {}}, + {"number": "not-an-int", "head": {"ref": MERGED_BRANCH}}, + ] + ) + self.assertEqual(index, {}) + self.assertEqual( + wca.resolve_owning_pr(branch=MERGED_BRANCH, pr_index=index)["status"], + wca.LINKAGE_NONE, + ) + + def test_detached_worktree_has_no_branch_linkage(self): + self.assertEqual( + wca.resolve_owning_pr(branch=None, pr_index={})["status"], + wca.LINKAGE_UNKNOWN, + ) + + +class TestUnrelatedClassificationsUnchanged(unittest.TestCase): + """Non-issue_work worktrees keep their pre-#858 classifications.""" + + PORCELAIN = _porcelain( + ("/repo", "master", "f" * 40), + ("/repo/branches/review-pr42", "review-pr42", "2" * 40), + ("/repo/branches/baseline-master-x", "baseline-master-x", "3" * 40), + ("/repo/branches/conflict-fix-pr50", "conflict-fix-pr50", "4" * 40), + ("/repo/branches/review-pr99", None, "5" * 40), + ) + + def _audit(self, **kwargs): + with patch.object( + wca, "list_worktrees", + return_value=wca.parse_worktree_porcelain(self.PORCELAIN), + ), patch.object( + wca, "read_worktree_dirty", + return_value={"exists": True, "dirty": False, "dirty_files": []}, + ), patch.object( + wca, "git_worktree_list", return_value="(mocked)" + ), patch.object( + wca, "is_head_ancestor_of_ref", return_value=True + ): + report = wca.audit_branches_directory("/repo", **kwargs) + return {wt["path"]: wt for wt in report["worktrees"]} + + def test_classifications_identical_with_and_without_pr_evidence(self): + without = self._audit() + with_evidence = self._audit( + pr_index=wca.build_pr_index([_pr(849, MERGED_BRANCH)]), + master_ref="prgs/master", + ) + self.assertEqual( + {p: e["classification"] for p, e in without.items()}, + {p: e["classification"] for p, e in with_evidence.items()}, + ) + + def test_lease_on_issue_does_not_capture_similarly_named_scratch_trees(self): + """A lease on issue 777 protects issue work, not baseline/review trees.""" + porcelain = _porcelain( + ("/repo/branches/baseline-master-issue-777", "baseline-issue-777", "7" * 40), + ("/repo/branches/issue-777-timeline", MERGED_BRANCH, HEAD_SHA), + ) + with patch.object( + wca, "list_worktrees", return_value=wca.parse_worktree_porcelain(porcelain) + ), patch.object( + wca, "read_worktree_dirty", + return_value={"exists": True, "dirty": False, "dirty_files": []}, + ), patch.object( + wca, "git_worktree_list", return_value="(mocked)" + ), patch.object( + wca, "is_head_ancestor_of_ref", return_value=True + ): + report = wca.audit_branches_directory( + "/repo", + pr_index=wca.build_pr_index([_pr(849, MERGED_BRANCH)]), + master_ref="prgs/master", + leased_issue_numbers={777}, + ) + by_path = {wt["path"]: wt for wt in report["worktrees"]} + + baseline = by_path["/repo/branches/baseline-master-issue-777"] + self.assertFalse(baseline["has_active_lease"]) + self.assertEqual(baseline["classification"], wca.CLASS_CLEAN_STALE_REMOVABLE) + + issue_work = by_path["/repo/branches/issue-777-timeline"] + self.assertTrue(issue_work["has_active_lease"]) + self.assertFalse(issue_work["removable"]) + + def test_review_and_baseline_still_removable(self): + by_path = self._audit( + pr_index=wca.build_pr_index([]), master_ref="prgs/master" + ) + self.assertEqual( + by_path["/repo/branches/review-pr42"]["classification"], + wca.CLASS_CLEAN_STALE_REMOVABLE, + ) + self.assertEqual( + by_path["/repo/branches/baseline-master-x"]["classification"], + wca.CLASS_CLEAN_STALE_REMOVABLE, + ) + self.assertEqual( + by_path["/repo/branches/review-pr99"]["classification"], + wca.CLASS_DETACHED_REVIEW_LEFTOVER, + ) + + def test_conflict_fix_ttl_behaviour_unchanged(self): + """conflict_fix still needs only TTL expiry; #858 did not touch it.""" + self.assertEqual( + wca.classify_worktree( + workflow_type=wca.WORKFLOW_CONFLICT_FIX, + is_dirty=False, + ttl_expired=True, + ), + wca.CLASS_CLEAN_STALE_REMOVABLE, + ) + self.assertEqual( + wca.classify_worktree( + workflow_type=wca.WORKFLOW_CONFLICT_FIX, + is_dirty=False, + ttl_expired=False, + ), + wca.CLASS_ACTIVE_ISSUE_WORK, + ) + + def test_issue_work_ttl_alone_no_longer_grants_removal(self): + """Age is not landing proof: TTL alone must not reclaim issue work.""" + self.assertEqual( + wca.classify_worktree( + workflow_type=wca.WORKFLOW_ISSUE_WORK, + is_dirty=False, + ttl_expired=True, + ), + wca.CLASS_ACTIVE_ISSUE_WORK, + ) + + +class TestAssessorPerformsNoDeletion(_AuditHarness): + def test_audit_never_removes_a_worktree(self): + with patch.object(wca, "remove_worktree") as removal: + self.merged_audit() + removal.assert_not_called() + + def test_audit_shells_out_to_no_destructive_git_command(self): + seen = [] + real_run = subprocess.run + + def recording_run(cmd, *args, **kwargs): + seen.append(cmd) + return real_run(["true"], *args, **kwargs) + + with patch.object(subprocess, "run", side_effect=recording_run): + wca.audit_branches_directory("/nonexistent-repo-for-audit") + + joined = [" ".join(c) if isinstance(c, list) else str(c) for c in seen] + for cmd in joined: + self.assertNotIn("worktree remove", cmd) + self.assertNotIn("branch -D", cmd) + self.assertNotIn("push", cmd) + + +class TestAgreementWithPrScopedReconciler(unittest.TestCase): + """The audit and merged_cleanup_reconcile must agree on identical input. + + Uses a real throwaway git repository so containment is computed by git + rather than asserted. Nothing outside the temporary directory is touched. + """ + + def _git(self, *args): + subprocess.run( + ["git", "-C", self.root, *args], + check=True, + capture_output=True, + text=True, + ) + + def setUp(self): + self._tmp = tempfile.TemporaryDirectory() + self.root = os.path.realpath(self._tmp.name) + self._git("init", "-b", "master", ".") + self._git("config", "user.email", "test@example.invalid") + self._git("config", "user.name", "Test") + with open(os.path.join(self.root, "seed.txt"), "w") as fh: + fh.write("seed\n") + self._git("add", "seed.txt") + self._git("commit", "-m", "seed") + + self.branch = "feat/issue-777-timeline" + self._git("checkout", "-b", self.branch) + with open(os.path.join(self.root, "feature.txt"), "w") as fh: + fh.write("feature\n") + self._git("add", "feature.txt") + self._git("commit", "-m", "feature") + self.head_sha = subprocess.run( + ["git", "-C", self.root, "rev-parse", "HEAD"], + capture_output=True, text=True, check=True, + ).stdout.strip() + self._git("checkout", "master") + self._git("merge", "--no-ff", "-m", "merge feature", self.branch) + + self.worktree = os.path.join(self.root, "branches", "issue-777-timeline") + self._git("worktree", "add", self.worktree, self.branch) + + def tearDown(self): + self._tmp.cleanup() + + def _pr_index(self): + return wca.build_pr_index( + [ + { + "number": 849, + "head": {"ref": self.branch, "sha": self.head_sha}, + "merged_at": "2026-07-24T01:00:00Z", + } + ] + ) + + def _audit_entry(self): + report = wca.audit_branches_directory( + self.root, pr_index=self._pr_index(), master_ref="master" + ) + return next(wt for wt in report["worktrees"] if wt["path"] == self.worktree) + + def _reconciler_entry(self): + return mcr.assess_local_worktree_cleanup( + pr_number=849, + head_branch=self.branch, + merged=True, + worktree_state=mcr.resolve_cleanup_worktree_state( + project_root=self.root, + head_branch=self.branch, + issue_number=777, + pr_head_sha=self.head_sha, + target_ref="master", + ), + active_lock=False, + ) + + def test_both_assessors_agree_the_worktree_is_safe(self): + audit_entry = self._audit_entry() + reconciler = self._reconciler_entry() + + self.assertTrue(reconciler["safe_to_remove_worktree"], reconciler) + self.assertTrue(audit_entry["removable"], audit_entry) + self.assertEqual(audit_entry["pr_number"], reconciler["pr_number"]) + self.assertEqual(audit_entry["merged_pr_cleanup"]["block_reasons"], []) + self.assertEqual(reconciler["block_reasons"], []) + + def test_both_assessors_agree_a_dirty_worktree_is_unsafe(self): + with open(os.path.join(self.worktree, "feature.txt"), "a") as fh: + fh.write("local edit\n") + + audit_entry = self._audit_entry() + reconciler = self._reconciler_entry() + + self.assertFalse(audit_entry["removable"]) + self.assertFalse(reconciler["safe_to_remove_worktree"]) + + def test_worktree_still_present_after_audit(self): + self._audit_entry() + self.assertTrue(os.path.isdir(self.worktree)) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/test_issue_871_durable_lock_head_refresh.py b/tests/test_issue_871_durable_lock_head_refresh.py new file mode 100644 index 0000000..37851b1 --- /dev/null +++ b/tests/test_issue_871_durable_lock_head_refresh.py @@ -0,0 +1,630 @@ +"""Durable linked-issue lock head refresh + merge-sync dead-session recovery (#871). + +``gitea_update_pr_branch_by_merge`` advances a PR's *remote* head but historically +never advanced the linked durable issue lock's recorded head. After the owning +session died the drifted lock became unrecoverable and no further synchronization +was possible (PR #866 / issue #855). + +Two halves are covered: + +* the write-side refresh (``issue_lock_store.assess/apply_durable_lock_head_refresh``) + that records the new synced head under compare-and-swap with read-after-write; and +* the read-side recovery relation (``issue_lock_recovery`` + + ``issue_lock_worktree.read_merge_sync_provenance``) that lets a dead-session lock + whose recorded head is a merge-sync *ancestor* of the live PR head be recovered — + and nothing else. +""" + +from __future__ import annotations + +import os +import subprocess +import sys +import tempfile +import unittest +from datetime import datetime, timedelta, timezone +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parent.parent)) + +import issue_lock_recovery # noqa: E402 +import issue_lock_store # noqa: E402 +import issue_lock_worktree # noqa: E402 + +ISSUE = 8710 +PR_NUMBER = 8711 +BRANCH = f"fix/issue-{ISSUE}-durable-lock-head-refresh" +IDENTITY = "example-user" +PROFILE = "example-author" +OLD = "a" * 40 +NEW1 = "b" * 40 +NEW2 = "c" * 40 +BASE = "d" * 40 +REMOTE = "prgs" +ORG = "ExampleOrg" +REPO = "ExampleRepo" + + +def dead_pid() -> int: + proc = subprocess.Popen([sys.executable, "-c", "pass"]) + proc.wait() + return proc.pid + + +def future_ts(hours: int = 4) -> str: + return ( + (datetime.now(timezone.utc) + timedelta(hours=hours)) + .isoformat() + .replace("+00:00", "Z") + ) + + +def _git(cwd, *args): + return subprocess.run( + ["git", "-C", cwd, *args], + capture_output=True, + text=True, + check=True, + ) + + +def _rev(cwd, ref="HEAD") -> str: + return _git(cwd, "rev-parse", ref).stdout.strip() + + +def build_merge_sync_repo(tmp: str) -> dict: + """Build a repo where a feature branch was synced by merging master in. + + Returns a dict with the prior (branch) head, the synced merge-commit head, + the master tip, plus a rebase-style linear descendant and an unrelated head. + """ + _git(tmp, "init", "-q", "-b", "master") + _git(tmp, "config", "user.email", "t@example.com") + _git(tmp, "config", "user.name", "T") + Path(tmp, "base.txt").write_text("base\n") + _git(tmp, "add", "-A") + _git(tmp, "commit", "-q", "-m", "root") + + # Feature branch cut from root, one commit — this is the PRIOR/recorded head. + _git(tmp, "checkout", "-q", "-b", BRANCH) + Path(tmp, "feature.txt").write_text("feature\n") + _git(tmp, "add", "-A") + _git(tmp, "commit", "-q", "-m", "feature work") + prior = _rev(tmp) + + # Master advances (the base the sync will merge in). + _git(tmp, "checkout", "-q", "master") + Path(tmp, "base.txt").write_text("base\nmore\n") + _git(tmp, "add", "-A") + _git(tmp, "commit", "-q", "-m", "master advance") + master_tip = _rev(tmp) + + # Sync: merge master INTO the feature branch → merge commit, first parent = prior. + _git(tmp, "checkout", "-q", BRANCH) + _git(tmp, "merge", "-q", "--no-ff", "-m", "Merge master into feature", "master") + synced = _rev(tmp) + + # A plain linear descendant of prior (NOT a merge) — a rebase/extra-commit shape. + _git(tmp, "checkout", "-q", "-b", "linear-branch", prior) + Path(tmp, "extra.txt").write_text("extra\n") + _git(tmp, "add", "-A") + _git(tmp, "commit", "-q", "-m", "extra linear commit") + linear = _rev(tmp) + + # An unrelated root (force-push / rewritten history shape). + unrelated_dir = tempfile.mkdtemp() + _git(unrelated_dir, "init", "-q", "-b", "x") + _git(unrelated_dir, "config", "user.email", "t@example.com") + _git(unrelated_dir, "config", "user.name", "T") + Path(unrelated_dir, "z.txt").write_text("z\n") + _git(unrelated_dir, "add", "-A") + _git(unrelated_dir, "commit", "-q", "-m", "unrelated") + unrelated = _rev(unrelated_dir) + + # Leave the worktree checked out on the feature branch at the PRIOR head, as + # a dead author session that never advanced would have left it. + _git(tmp, "checkout", "-q", BRANCH) + _git(tmp, "reset", "-q", "--hard", prior) + + return { + "prior": prior, + "master_tip": master_tip, + "synced": synced, + "linear": linear, + "unrelated": unrelated, + } + + +# ─────────────────────────── write-side refresh ─────────────────────────── + + +class TestDurableLockHeadRefresh(unittest.TestCase): + def setUp(self): + self.lock_dir = tempfile.mkdtemp() + self.wt = tempfile.mkdtemp() + lock_data = { + "issue_number": ISSUE, + "branch_name": BRANCH, + "worktree_path": self.wt, + "remote": REMOTE, + "org": ORG, + "repo": REPO, + "claimant": {"username": IDENTITY, "profile": PROFILE}, + "work_lease": { + "operation_type": issue_lock_store.AUTHOR_ISSUE_WORK_LEASE, + "issue_number": ISSUE, + "branch": BRANCH, + "worktree_path": self.wt, + "claimant": {"username": IDENTITY, "profile": PROFILE}, + "expires_at": future_ts(), + }, + } + issue_lock_store.bind_session_lock(lock_data, lock_dir=self.lock_dir) + + def _apply(self, **over): + kw = dict( + remote=REMOTE, org=ORG, repo=REPO, issue_number=ISSUE, + branch_name=BRANCH, worktree_path=self.wt, pr_number=PR_NUMBER, + identity=IDENTITY, profile=PROFILE, current_pid=os.getpid(), + expected_old_head=OLD, new_head=NEW1, synced_at=future_ts(0), + base_head=BASE, lock_dir=self.lock_dir, + ) + kw.update(over) + return issue_lock_store.apply_durable_lock_head_refresh(**kw) + + def _load(self): + return issue_lock_store.load_issue_lock( + remote=REMOTE, org=ORG, repo=REPO, issue_number=ISSUE, + lock_dir=self.lock_dir, + ) + + def test_first_sync_updates_recorded_head(self): + """AC1: first base sync writes the resulting head to the durable lock.""" + res = self._apply() + self.assertTrue(res["refreshed"], res["reasons"]) + self.assertTrue(res["read_after_write_ok"]) + self.assertEqual(self._load().get("synced_pr_head"), NEW1) + + def test_second_sync_after_master_advance(self): + """AC2: a later master advance permits a second sanctioned sync.""" + self.assertTrue(self._apply()["refreshed"]) + res2 = self._apply(expected_old_head=NEW1, new_head=NEW2) + self.assertTrue(res2["refreshed"], res2["reasons"]) + self.assertEqual(self._load().get("synced_pr_head"), NEW2) + history = self._load().get("branch_sync_history") + self.assertEqual(len(history), 2) + self.assertEqual(history[0]["last_synced_pr_head"], NEW1) + self.assertEqual(history[1]["prior_pr_head"], NEW1) + + def test_cas_detects_concurrent_head_change(self): + """AC6: CAS refuses when the recorded synced head is not the old head.""" + self.assertTrue(self._apply()["refreshed"]) # recorded head now NEW1 + # A second sync claiming the old head is still OLD must fail closed. + res = self._apply(expected_old_head=OLD, new_head=NEW2) + self.assertFalse(res["refreshed"]) + self.assertTrue(any("CAS" in r or "concurrent" in r for r in res["reasons"])) + self.assertEqual(self._load().get("synced_pr_head"), NEW1) + + def test_wrong_issue_fails_closed(self): + res = self._apply(issue_number=999999) + self.assertFalse(res["refreshed"]) + + def test_wrong_branch_fails_closed(self): + res = self._apply(branch_name="fix/issue-8710-wrong") + self.assertFalse(res["refreshed"]) + + def test_wrong_repo_fails_closed(self): + res = self._apply(repo="OtherRepo") + self.assertFalse(res["refreshed"]) + + def test_wrong_identity_fails_closed(self): + res = self._apply(identity="intruder") + self.assertFalse(res["refreshed"]) + + def test_wrong_profile_fails_closed(self): + res = self._apply(profile="prgs-reviewer") + self.assertFalse(res["refreshed"]) + + def test_foreign_session_fails_closed(self): + """A refresh is not a recovery: the current process must own the lock.""" + path = issue_lock_store.lock_file_path( + remote=REMOTE, org=ORG, repo=REPO, issue_number=ISSUE, + lock_dir=self.lock_dir, + ) + rec = issue_lock_store.read_lock_file(path) + rec["session_pid"] = dead_pid() + rec["pid"] = rec["session_pid"] + issue_lock_store.save_lock_file(path, rec) + res = self._apply() + self.assertFalse(res["refreshed"]) + self.assertTrue(any("current session" in r or "live owner" in r for r in res["reasons"])) + + def test_new_equals_old_fails_closed(self): + res = self._apply(expected_old_head=OLD, new_head=OLD) + self.assertFalse(res["refreshed"]) + + def test_non_full_sha_fails_closed(self): + self.assertFalse(self._apply(new_head="deadbeef")["refreshed"]) + self.assertFalse(self._apply(expected_old_head="xyz")["refreshed"]) + + def test_no_lock_fails_closed(self): + assessment = issue_lock_store.assess_durable_lock_head_refresh( + None, remote=REMOTE, org=ORG, repo=REPO, issue_number=ISSUE, + branch_name=BRANCH, worktree_path=self.wt, pr_number=PR_NUMBER, + identity=IDENTITY, profile=PROFILE, current_pid=os.getpid(), + expected_old_head=OLD, new_head=NEW1, + ) + self.assertFalse(assessment["allowed"]) + + +# ─────────────────────── merge-sync provenance (real git) ─────────────────── + + +class TestMergeSyncProvenanceObservation(unittest.TestCase): + def setUp(self): + self.tmp = tempfile.mkdtemp() + self.shas = build_merge_sync_repo(self.tmp) + + def test_merge_sync_is_recognized(self): + obs = issue_lock_worktree.read_merge_sync_provenance( + self.tmp, prior_head_sha=self.shas["prior"], + synced_head_sha=self.shas["synced"], + ) + self.assertTrue(obs["is_merge_sync"], obs["reasons"]) + self.assertTrue(obs["prior_is_ancestor"]) + self.assertTrue(obs["synced_is_merge"]) + self.assertTrue(obs["first_parent_reaches_prior"]) + + def test_linear_descendant_is_not_a_merge_sync(self): + """A plain non-merge descendant (rebase/extra commit) is not a sync.""" + obs = issue_lock_worktree.read_merge_sync_provenance( + self.tmp, prior_head_sha=self.shas["prior"], + synced_head_sha=self.shas["linear"], + ) + self.assertTrue(obs["probe_ok"]) + self.assertFalse(obs["is_merge_sync"]) + self.assertFalse(obs["synced_is_merge"]) + + def test_unrelated_history_fails_closed(self): + """A rewritten/force-pushed head where prior is unreachable fails closed.""" + obs = issue_lock_worktree.read_merge_sync_provenance( + self.tmp, prior_head_sha=self.shas["prior"], + synced_head_sha=self.shas["unrelated"], + ) + self.assertFalse(obs["is_merge_sync"]) + + def test_missing_args_fail_closed(self): + obs = issue_lock_worktree.read_merge_sync_provenance( + self.tmp, prior_head_sha=None, synced_head_sha=self.shas["synced"], + ) + self.assertFalse(obs["is_merge_sync"]) + + +# ──────────────────── merge-sync dead-session recovery ────────────────────── + + +def make_dead_lock(worktree, **over): + pid = dead_pid() + lock = { + "issue_number": ISSUE, + "branch_name": BRANCH, + "worktree_path": worktree, + "remote": REMOTE, + "org": ORG, + "repo": REPO, + "session_pid": pid, + "pid": pid, + "claimant": {"username": IDENTITY, "profile": PROFILE}, + "work_lease": { + "operation_type": issue_lock_store.AUTHOR_ISSUE_WORK_LEASE, + "issue_number": ISSUE, + "branch": BRANCH, + "worktree_path": worktree, + "claimant": {"username": IDENTITY, "profile": PROFILE}, + "expires_at": future_ts(), + }, + } + lock.update(over) + return lock + + +def sync_prov(prior, synced, **over): + d = { + "prior_head_sha": prior, + "synced_head_sha": synced, + "probe_ok": True, + "prior_present": True, + "synced_present": True, + "prior_is_ancestor": True, + "synced_is_merge": True, + "first_parent_reaches_prior": True, + "is_merge_sync": True, + "first_parent_sha": prior, + "parent_count": 2, + "proof": f"{synced} merged base into branch above {prior}", + "reasons": [], + } + d.update(over) + return d + + +class TestMergeSyncRecovery(unittest.TestCase): + def setUp(self): + self.tmp = tempfile.mkdtemp() + self.shas = build_merge_sync_repo(self.tmp) + self.prior = self.shas["prior"] + self.synced = self.shas["synced"] + + def _assess(self, **over): + lock = over.pop("_lock", None) or make_dead_lock(self.tmp) + kw = dict( + issue_number=ISSUE, branch_name=BRANCH, worktree_path=self.tmp, + remote=REMOTE, org=ORG, repo=REPO, identity=IDENTITY, profile=PROFILE, + current_branch=BRANCH, porcelain_status="", + head_sha=self.prior, remote_head_sha=self.synced, + pr_head_sha=self.synced, pr_number=PR_NUMBER, + competing_live_locks=[], candidate_branches=[BRANCH], + current_pid=os.getpid(), + remote_branch_exists=True, + sync_provenance=sync_prov(self.prior, self.synced), + ) + kw.update(over) + return issue_lock_recovery.assess_dead_session_lock_recovery(lock, **kw) + + def test_merge_sync_drift_is_recoverable(self): + """AC3/AC4: dead session, recorded head is a merge-sync ancestor of PR head.""" + res = self._assess() + self.assertEqual(res["outcome"], issue_lock_recovery.RECOVERY_SANCTIONED, res["reasons"]) + self.assertEqual( + res["evidence"]["head_relation"], + issue_lock_recovery.HEAD_RELATION_REMOTE_MERGE_SYNCED, + ) + self.assertEqual(res["evidence"]["accepted_head"], self.synced) + + def test_missing_provenance_fails_closed(self): + """No server-derived provenance → cannot accept a remote ahead of local.""" + res = self._assess(sync_provenance=None) + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_non_ancestor_recorded_head_fails_closed(self): + """AC7: provenance that does not prove ancestry is rejected.""" + res = self._assess( + sync_provenance=sync_prov( + self.prior, self.synced, prior_is_ancestor=False, is_merge_sync=False, + reasons=["prior head is not an ancestor"], + ) + ) + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_force_pushed_history_fails_closed(self): + """AC8: a rewritten head (not a merge sync) stays protected.""" + res = self._assess( + sync_provenance=sync_prov( + self.prior, self.synced, is_merge_sync=False, synced_is_merge=False, + reasons=["not a merge-based sync"], + ) + ) + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_provenance_for_other_commits_fails_closed(self): + """Provenance whose endpoints differ from the heads under assessment is rejected.""" + res = self._assess( + sync_provenance=sync_prov("f" * 40, self.synced), + ) + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_dirty_worktree_fails_closed(self): + """AC11: dirty worktrees remain protected.""" + res = self._assess(porcelain_status=" M feature.txt\n") + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_live_owner_fails_closed(self): + """AC10: a live recorded owner is not a dead-session recovery.""" + lock = make_dead_lock(self.tmp, session_pid=os.getpid(), pid=os.getpid()) + res = self._assess(_lock=lock) + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_competing_claimant_fails_closed(self): + """AC13: a competing live lock blocks recovery.""" + res = self._assess( + competing_live_locks=[{ + "issue_number": ISSUE, "branch_name": BRANCH, + "worktree_path": "/some/other/wt", "pid": os.getpid(), + }] + ) + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_wrong_branch_fails_closed(self): + """AC9: worktree on a different branch fails closed.""" + res = self._assess(current_branch="fix/issue-8710-other") + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_wrong_identity_fails_closed(self): + res = self._assess(identity="intruder") + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_pr_head_mismatch_fails_closed(self): + """The open PR must sit at the synced remote head.""" + res = self._assess(pr_head_sha="e" * 40) + self.assertEqual(res["outcome"], issue_lock_recovery.REFUSED) + + def test_owning_pr_evidence_for_merge_sync(self): + res = self._assess() + ev = issue_lock_recovery.owning_pr_recovery_evidence(res) + self.assertIsNotNone(ev) + self.assertEqual(ev["pr_number"], PR_NUMBER) + self.assertEqual(ev["head_sha"], self.synced) + self.assertEqual( + ev["head_relation"], + issue_lock_recovery.HEAD_RELATION_REMOTE_MERGE_SYNCED, + ) + + def test_recovered_owning_pr_from_persisted_record(self): + res = self._assess() + record = issue_lock_recovery.build_recovery_record(res, recovered_at=future_ts(0)) + lock = {"issue_number": ISSUE, "branch_name": BRANCH, + "dead_session_recovery": record} + rebuilt = issue_lock_recovery.recovered_owning_pr_from_lock(lock) + self.assertIsNotNone(rebuilt) + self.assertEqual(rebuilt["head_sha"], self.synced) + self.assertEqual( + rebuilt["head_relation"], + issue_lock_recovery.HEAD_RELATION_REMOTE_MERGE_SYNCED, + ) + + +class TestExistingRelationsUnchanged(unittest.TestCase): + """AC14/AC15: equal-head recovery still works; merge-sync did not weaken it.""" + + def setUp(self): + self.tmp = tempfile.mkdtemp() + self.shas = build_merge_sync_repo(self.tmp) + + def test_equal_head_recovery_still_sanctioned(self): + # Worktree at prior head; remote also at prior head → the #753 equal case. + prior = self.shas["prior"] + lock = make_dead_lock(self.tmp) + res = issue_lock_recovery.assess_dead_session_lock_recovery( + lock, issue_number=ISSUE, branch_name=BRANCH, worktree_path=self.tmp, + remote=REMOTE, org=ORG, repo=REPO, identity=IDENTITY, profile=PROFILE, + current_branch=BRANCH, porcelain_status="", + head_sha=prior, remote_head_sha=prior, + pr_head_sha=prior, pr_number=PR_NUMBER, + competing_live_locks=[], candidate_branches=[BRANCH], + current_pid=os.getpid(), remote_branch_exists=True, + ) + self.assertEqual(res["outcome"], issue_lock_recovery.RECOVERY_SANCTIONED, res["reasons"]) + self.assertEqual( + res["evidence"]["head_relation"], issue_lock_recovery.HEAD_RELATION_EQUAL, + ) + + +class TestUpdatePrWrapperPartialFailure(unittest.TestCase): + """AC5/AC16: the tool advances the remote head then refreshes the durable lock. + + When the durable refresh fails after the remote advance, the tool must report a + partial lifecycle failure and NOT a fully successful synchronization. Exact PR- + head / base-head pinning is preserved (delegated to the real preflight, stubbed + here only to isolate the post-update lifecycle branch). + """ + + def setUp(self): + import gitea_mcp_server as gms # noqa: E402 + self.gms = gms + self._orig = {} + + def _patch(name, value): + self._orig[name] = getattr(gms, name) + setattr(gms, name, value) + + _patch("get_profile", lambda *a, **k: { + "allowed_operations": ["gitea.branch.push"], + "forbidden_operations": [], + "profile_name": "prgs-author", + }) + _patch("_role_kind", lambda *a, **k: "author") + _patch("_profile_operation_gate", lambda *a, **k: None) + _patch("_permission_block_report", lambda *a, **k: {}) + _patch("_resolve", lambda *a, **k: ("gitea.prgs.cc", ORG, REPO)) + _patch("_verify_role_mutation_workspace", lambda *a, **k: None) + _patch("_get_workspace_porcelain", lambda *a, **k: "") + _patch("_canonical_local_git_root", lambda *a, **k: "/x") + _patch("_master_parity_block", lambda *a, **k: None) + _patch("_auth", lambda *a, **k: {"token": "x"}) + _patch("repo_api_url", lambda *a, **k: "http://api") + _patch("_redact", lambda s: s) + _patch("_work_lease_claimant", lambda *a, **k: { + "username": IDENTITY, "profile": PROFILE, + }) + _patch("_prove_author_ownership_for_pr", lambda *a, **k: { + "has_author_lock": True, "matched_issue": ISSUE, + "matched_via": "branch", "linked_issues": [ISSUE], + "recovered_owning_pr": None, "reasons": [], + }) + + # Real preflight is unit-tested elsewhere; stub it to isolate the + # post-update durable-lock lifecycle branch under test. + orig_pf = gms.pr_sync_status.assess_update_pr_branch_preflight + self._orig_pf = orig_pf + gms.pr_sync_status.assess_update_pr_branch_preflight = ( + lambda *a, **k: {"mutation_allowed": True, "reasons": [], "performed": False} + ) + + # Sequence the two GET /pulls calls: OLD before update, NEW after. + self._pull_calls = {"n": 0} + + def fake_api_request(method, url, auth, *a, **k): + m = method.upper() + if m == "GET" and url.endswith(f"/pulls/{PR_NUMBER}"): + self._pull_calls["n"] += 1 + head = OLD if self._pull_calls["n"] == 1 else NEW1 + return { + "state": "open", + "head": {"sha": head, "ref": BRANCH}, + "base": {"sha": BASE, "ref": "master"}, + "mergeable": True, "title": "t", "body": "b", + } + if m == "GET" and "/branches/" in url: + return {"commit": {"id": BASE}} + if m == "POST" and "/update" in url: + return {} + return {} + + _patch("api_request", fake_api_request) + + def tearDown(self): + for name, value in self._orig.items(): + setattr(self.gms, name, value) + self.gms.pr_sync_status.assess_update_pr_branch_preflight = self._orig_pf + + def _run(self): + return self.gms.gitea_update_pr_branch_by_merge( + pr_number=PR_NUMBER, + expected_pr_head_sha=OLD, + expected_base_head_sha=BASE, + remote=REMOTE, + worktree_path="/tmp/branches/wt-871", + ) + + def test_partial_failure_when_refresh_fails(self): + self._orig["apply_durable_lock_head_refresh"] = ( + self.gms.issue_lock_store.apply_durable_lock_head_refresh + ) + self.gms.issue_lock_store.apply_durable_lock_head_refresh = ( + lambda **k: {"refreshed": False, "reasons": ["forced refresh failure"]} + ) + try: + res = self._run() + finally: + self.gms.issue_lock_store.apply_durable_lock_head_refresh = ( + self._orig["apply_durable_lock_head_refresh"] + ) + self.assertTrue(res["performed"]) + self.assertEqual(res["new_pr_head_sha"], NEW1) + self.assertFalse(res["success"]) + self.assertTrue(res["partial_lifecycle_failure"]) + self.assertFalse(res["durable_lock_refreshed"]) + + def test_full_success_when_refresh_succeeds(self): + self._orig["apply_durable_lock_head_refresh"] = ( + self.gms.issue_lock_store.apply_durable_lock_head_refresh + ) + self.gms.issue_lock_store.apply_durable_lock_head_refresh = ( + lambda **k: {"refreshed": True, "read_after_write_ok": True, + "new_head": NEW1, "reasons": ["ok"]} + ) + try: + res = self._run() + finally: + self.gms.issue_lock_store.apply_durable_lock_head_refresh = ( + self._orig["apply_durable_lock_head_refresh"] + ) + self.assertTrue(res["success"]) + self.assertTrue(res["performed"]) + self.assertTrue(res["durable_lock_refreshed"]) + self.assertTrue(res["fully_synchronized"]) + self.assertEqual(res["new_pr_head_sha"], NEW1) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/test_issue_lock_store.py b/tests/test_issue_lock_store.py index 9c41907..a8a4a7b 100644 --- a/tests/test_issue_lock_store.py +++ b/tests/test_issue_lock_store.py @@ -24,6 +24,8 @@ def _lease(expires_at: str) -> dict: def _lock_record(**overrides) -> dict: + # #860: live locks require a usable session pid; PID-less records are never + # classified live merely because expiry/heartbeat fields are present. record = { "issue_number": 420, "branch_name": "feat/issue-420-server-code-parity", @@ -31,6 +33,8 @@ def _lock_record(**overrides) -> dict: "org": "Scaled-Tech-Consulting", "repo": "Gitea-Tools", "worktree_path": "/tmp/wt-420", + "session_pid": os.getpid(), + "pid": os.getpid(), "work_lease": _lease("2999-01-01T00:00:00Z"), } record.update(overrides) @@ -88,6 +92,8 @@ class TestIssueLockStore(unittest.TestCase): existing = _lock_record( branch_name="feat/issue-420-other", worktree_path="/tmp/other", + session_pid=os.getpid(), + pid=os.getpid(), work_lease=_lease("2999-01-01T00:00:00Z"), ) path = ils.lock_file_path( diff --git a/tests/test_mcp_restart_governance_docs.py b/tests/test_mcp_restart_governance_docs.py new file mode 100644 index 0000000..5b12e50 --- /dev/null +++ b/tests/test_mcp_restart_governance_docs.py @@ -0,0 +1,107 @@ +"""Documentation acceptance for the MCP restart governance ADR (#656). + +Enforces issue #656 acceptance criteria: + +* AC1 — policy document exists with an authorization matrix and the recorded + v1 decision (controller approval + automated safety gates). +* AC2 — restart is stated as a last resort with enumerated narrower recoveries. +* AC3 — a unilateral LLM full restart with affected sessions is forbidden. +* AC4 — break-glass conditions are listed. +* AC5 — the ADR is linked to #655, #652, #653, #630, #642, and is cross-linked + from the safety model and the web-console deployment boundary docs. +""" +from pathlib import Path + +REPO_ROOT = Path(__file__).resolve().parent.parent +ADR = REPO_ROOT / "docs" / "architecture" / "mcp-restart-governance.md" +ADR_BASENAME = "mcp-restart-governance.md" + +CROSS_LINK_DOCS = ( + REPO_ROOT / "docs" / "safety-model.md", + REPO_ROOT / "docs" / "webui-deployment.md", +) + +LINKED_ISSUES = ("#655", "#652", "#653", "#630", "#642") +POLICY_IDS = ("RG-01", "RG-02", "RG-03", "RG-04", "RG-05", "RG-06", "RG-07", "RG-08") + + +def _read(path: Path) -> str: + assert path.is_file(), f"missing {path.relative_to(REPO_ROOT)}" + return path.read_text(encoding="utf-8") + + +def test_ac1_adr_exists_with_matrix_and_v1_decision(): + text = _read(ADR) + lower = text.lower() + assert text.lstrip().startswith("#"), "ADR lacks a title" + assert "#656" in text + assert "authorization matrix" in lower + # The matrix is a real table with the worker and privileged roles. + for role in ("author", "reviewer", "merger", "reconciler", "controller", + "operator", "admin"): + assert role in lower, f"authorization matrix missing role {role!r}" + # Recorded v1 decision. + assert "restart-governance/v1" in text + assert "controller approval" in lower and "automated safety gates" in lower + + +def test_ac2_restart_is_last_resort_with_narrower_recoveries(): + text = _read(ADR) + lower = text.lower() + assert "last resort" in lower + # Enumerated narrower recoveries precede full restart on the ladder. + for rung in ("reconnect", "rebind", "scoped restart", "full restart", + "host"): + assert rung in lower, f"recovery ladder missing rung {rung!r}" + + +def test_ac3_forbids_unilateral_llm_full_restart_with_affected_sessions(): + text = _read(ADR) + lower = text.lower() + assert "forbidden" in lower + assert "llm" in lower and "restart" in lower + assert "unilateral" in lower + # A worker role must not perform or authorize full/host restart. + assert "must not" in lower + + +def test_ac4_break_glass_conditions_listed(): + text = _read(ADR) + lower = text.lower() + assert "break-glass" in lower + assert "incident" in lower + assert "audit" in lower + + +def test_ac5_adr_links_issue_lineage(): + text = _read(ADR) + for issue in LINKED_ISSUES: + assert issue in text, f"ADR must link issue {issue}" + + +def test_ac5_safety_model_and_deployment_cross_link_adr(): + for path in CROSS_LINK_DOCS: + text = _read(path) + assert ADR_BASENAME in text, ( + f"{path.relative_to(REPO_ROOT)} must cross-link {ADR_BASENAME} " + f"(issue #656 acceptance criterion 5)" + ) + + +def test_policy_ids_present_for_enforcement_code(): + text = _read(ADR) + for pid in POLICY_IDS: + assert pid in text, f"policy id {pid} missing from ADR" + + +def test_failure_behavior_denies_on_ambiguity(): + text = _read(ADR) + lower = text.lower() + assert "ambiguous" in lower and "deny" in lower + + +def test_cross_links_do_not_embed_secrets(): + for path in (ADR,) + CROSS_LINK_DOCS: + text = _read(path) + for marker in ("ghp_", "BEGIN PRIVATE KEY", "Authorization: Bearer"): + assert marker not in text, f"{path} contains {marker!r}" diff --git a/tests/test_mcp_restart_paths.py b/tests/test_mcp_restart_paths.py new file mode 100644 index 0000000..cf0c5eb --- /dev/null +++ b/tests/test_mcp_restart_paths.py @@ -0,0 +1,146 @@ +"""Tests for the MCP restart-path inventory and guards (#657). + +Covers: +* the registry is well-formed and every path is classified; +* unknown restart attempts fail closed (AC "fail closed on unknown restart"); +* the previously-unguarded full-restart primitives stay guarded/absent + against the real source tree (AC "tests for at least one previously + unguarded path"); +* pkill of the daemon is still classified as contamination (#630, AC3); +* the inventory doc and module stay in lock-step. +""" + +import os +import tempfile +import unittest +from pathlib import Path + +import mcp_restart_paths as rp +import runtime_recovery_guard + +REPO_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__))) +DOC_PATH = os.path.join(REPO_ROOT, "docs", "mcp-restart-path-inventory.md") + + +class TestRegistryWellformed(unittest.TestCase): + def test_registry_is_wellformed(self): + # Must not raise. + rp.assert_registry_wellformed() + + def test_every_path_has_valid_classification(self): + for path in rp.iter_restart_paths(): + self.assertIn(path.classification, rp.VALID_CLASSIFICATIONS) + self.assertTrue(path.guard.strip(), path.path_id) + self.assertTrue(path.references, path.path_id) + self.assertTrue(path.locations, path.path_id) + + def test_ids_are_unique(self): + ids = [p.path_id for p in rp.iter_restart_paths()] + self.assertEqual(len(ids), len(set(ids))) + + def test_covers_every_classification(self): + present = {p.classification for p in rp.iter_restart_paths()} + self.assertEqual(present, set(rp.VALID_CLASSIFICATIONS)) + + +class TestUnknownAttemptFailsClosed(unittest.TestCase): + def test_unknown_path_raises(self): + with self.assertRaises(rp.UnknownRestartPathError): + rp.assert_restart_attempt_registered("totally_novel_restart_hack") + + def test_get_unknown_raises(self): + with self.assertRaises(rp.UnknownRestartPathError): + rp.get_restart_path("nope") + + def test_registered_attempt_returns_path(self): + path = rp.assert_restart_attempt_registered("manual_daemon_kill") + self.assertEqual(path.classification, rp.CLASS_FORBIDDEN) + + +class TestDaemonNeverSelfReplaces(unittest.TestCase): + """Previously-unguarded full-restart primitive: daemon self-replacement.""" + + def test_no_self_replacement_in_source(self): + # The live daemon modules must contain no os.execv/os.kill/os._exit + # self-restart call. Must not raise. + rp.assert_no_daemon_self_replacement(REPO_ROOT) + + def test_scanner_flags_injected_violation(self): + # Guard the guard: prove the scanner catches a real self-replace call. + with tempfile.TemporaryDirectory() as tmp: + bad = Path(tmp) / "gitea_mcp_server.py" + bad.write_text( + "import os\n" + "def restart():\n" + " os.execv('/usr/bin/python', ['python'])\n", + encoding="utf-8", + ) + found = rp.scan_daemon_self_replacement(tmp) + self.assertTrue(found) + with self.assertRaises(AssertionError): + rp.assert_no_daemon_self_replacement(tmp) + + def test_scanner_ignores_comment_and_docstring_mentions(self): + with tempfile.TemporaryDirectory() as tmp: + ok = Path(tmp) / "gitea_mcp_server.py" + ok.write_text( + "import os\n" + "# NOT os.execv() to re-point the interpreter here.\n" + '"""Never calls os._exit to restart."""\n' + "value = 1\n", + encoding="utf-8", + ) + self.assertEqual(rp.scan_daemon_self_replacement(tmp), []) + + +class TestLegacyAutoRestartHelperRemoved(unittest.TestCase): + """Previously-unguarded full-restart path: _trigger_mcp_auto_restart.""" + + def test_helper_absent_in_source(self): + # Must not raise: helper was removed in #685. + rp.assert_auto_restart_helper_absent(REPO_ROOT) + + def test_scanner_flags_reintroduced_helper(self): + with tempfile.TemporaryDirectory() as tmp: + bad = Path(tmp) / "mcp_server.py" + bad.write_text( + "def _trigger_mcp_auto_restart():\n return True\n", + encoding="utf-8", + ) + with self.assertRaises(AssertionError): + rp.assert_auto_restart_helper_absent(tmp) + + +class TestPkillStaysForbidden(unittest.TestCase): + """AC3: pkill of the daemon remains forbidden/contaminating (#630).""" + + def test_manual_daemon_kill_registered_as_forbidden(self): + path = rp.get_restart_path("manual_daemon_kill") + self.assertEqual(path.classification, rp.CLASS_FORBIDDEN) + + def test_pkill_classified_as_contamination(self): + assessment = runtime_recovery_guard.assess_recovery_command( + "pkill -f mcp_server.py" + ) + self.assertTrue(assessment["contaminated"]) + + def test_read_only_probe_not_contamination(self): + assessment = runtime_recovery_guard.assess_recovery_command( + "ps aux | grep mcp_server" + ) + self.assertFalse(assessment["contaminated"]) + + +class TestInventoryDocInSync(unittest.TestCase): + def test_doc_exists(self): + self.assertTrue(os.path.exists(DOC_PATH), DOC_PATH) + + def test_doc_mentions_every_path_id(self): + with open(DOC_PATH, encoding="utf-8") as handle: + doc = handle.read() + for path in rp.iter_restart_paths(): + self.assertIn(path.path_id, doc, f"doc missing {path.path_id}") + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/test_merged_cleanup_reconcile.py b/tests/test_merged_cleanup_reconcile.py index 284ad37..e9a8709 100644 --- a/tests/test_merged_cleanup_reconcile.py +++ b/tests/test_merged_cleanup_reconcile.py @@ -12,6 +12,59 @@ import merged_cleanup_reconcile as mcr # noqa: E402 class TestMergedCleanupAssessment(unittest.TestCase): + def test_issue_851_plan_order_worktree_then_reassess_then_remote(self): + """#851 dry-run plan: remove worktree, reassess ownership, then remote.""" + plan = mcr.plan_cleanup_execution_order( + remote_assessment={"safe_to_delete_remote": True}, + local_assessment={"safe_to_remove_worktree": True}, + ) + actions = [s["action"] for s in plan] + self.assertEqual( + actions, + [ + "remove_local_worktree", + "reassess_branch_ownership", + "delete_remote_branch", + ], + ) + self.assertEqual(plan[0]["phase"], 1) + self.assertEqual(plan[-1]["phase"], 3) + self.assertIn("independently_safe", plan[0]["reason"]) + self.assertIn("reassessment", plan[-1]["reason"]) + + def test_issue_851_plan_remote_only_when_worktree_not_safe(self): + plan = mcr.plan_cleanup_execution_order( + remote_assessment={"safe_to_delete_remote": True}, + local_assessment={"safe_to_remove_worktree": False}, + ) + self.assertEqual([s["action"] for s in plan], ["delete_remote_branch"]) + self.assertNotIn("reassess_branch_ownership", [s["action"] for s in plan]) + + def test_issue_851_plan_worktree_only_when_remote_not_safe(self): + plan = mcr.plan_cleanup_execution_order( + remote_assessment={"safe_to_delete_remote": False}, + local_assessment={"safe_to_remove_worktree": True}, + ) + self.assertEqual([s["action"] for s in plan], ["remove_local_worktree"]) + + def test_issue_851_entry_includes_planned_execution_order(self): + entry = mcr.build_pr_cleanup_entry( + pr={ + "number": 848, + "title": "Closes #844", + "body": "", + "merged_at": "2026-07-23T00:00:00Z", + "head": {"ref": "fix/issue-844-x", "sha": "a" * 40}, + }, + project_root="/tmp/not-a-real-root", + open_pr_heads=set(), + remote_branch_exists=True, + head_on_master=True, + delete_capability_allowed=True, + ) + self.assertIn("planned_execution_order", entry) + self.assertIsInstance(entry["planned_execution_order"], list) + def test_extract_linked_issue_from_closes(self): issue = mcr.extract_linked_issue( "feat: cleanup (Closes #269)", diff --git a/tests/test_pr_ownership_issue_pr_mismatch.py b/tests/test_pr_ownership_issue_pr_mismatch.py index 2286dae..99b500c 100644 --- a/tests/test_pr_ownership_issue_pr_mismatch.py +++ b/tests/test_pr_ownership_issue_pr_mismatch.py @@ -37,6 +37,7 @@ def _live_lock( "operation_type": issue_lock_store.AUTHOR_ISSUE_WORK_LEASE, "acquired_at": now.isoformat(), "expires_at": (now + timedelta(hours=2)).isoformat(), + "session_pid": os.getpid(), "owner_pid": os.getpid(), "status": "active", } @@ -177,11 +178,24 @@ class TestAuthorOwnershipIssuePrMismatch(unittest.TestCase): self.assertFalse(result["proven"], result) self.assertTrue(any("branch" in r for r in result["reasons"])) - def test_no_lock_fail_closed(self): + def test_pidless_durable_lock_rejected(self): + """A lock without any PID identity must be classified as malformed/non-live and fail closed.""" + lock = _live_lock(issue_number=727) + lock.pop("session_pid", None) + lock.pop("owner_pid", None) + lock.pop("pid", None) + path = issue_lock_store.lock_file_path( + remote="prgs", + org="Scaled-Tech-Consulting", + repo="Gitea-Tools", + issue_number=727, + lock_dir=self.lock_dir, + ) + issue_lock_store.save_lock_file(path, lock) result = mcp._prove_author_ownership_for_pr( pr_number=728, pr_title="feat: pr sync", - pr_body="Closes #727", + pr_body="Fixes #727", source_branch="feat/issue-727-pr-sync-status", remote="prgs", host=None, diff --git a/tests/test_pr_work_lease.py b/tests/test_pr_work_lease.py index 8dbcb33..ddbf2cc 100644 --- a/tests/test_pr_work_lease.py +++ b/tests/test_pr_work_lease.py @@ -19,6 +19,7 @@ from pr_work_lease import ( # noqa: E402 assess_reviewer_mutation_blocked, assess_reviewer_stale_head_final_report, format_conflict_fix_lease_body, + find_active_conflict_fix_lease, parse_conflict_fix_lease_comment, parse_reviewer_lease_comment, ) @@ -203,5 +204,157 @@ class TestFormatLease(unittest.TestCase): self.assertEqual(parsed["pr_number"], 376) +class TestConflictFixLeaseLifecycle(unittest.TestCase): + def test_claim_followed_by_matching_release(self): + claim_body = _conflict_fix_body(phase="claimed", worktree="branches/fix-376") + expires = (NOW + timedelta(minutes=60)).isoformat().replace("+00:00", "Z") + release_body = "\n".join([ + CONFLICT_FIX_LEASE_MARKER, + "pr: #376", + "branch: feat/fix-376", + "worktree: branches/fix-376", + "profile: prgs-author", + "phase: released", + f"head_before: {HEAD_A}", + f"head_after: {HEAD_B}", + f"expires_at: {expires}", + ]) + comments = [{"body": claim_body}, {"body": release_body}] + lease = find_active_conflict_fix_lease(comments, pr_number=376, now=NOW) + self.assertIsNone(lease) + + def test_expired_claim_without_release(self): + past_expires = (NOW - timedelta(minutes=10)).isoformat().replace("+00:00", "Z") + claim_body = "\n".join([ + CONFLICT_FIX_LEASE_MARKER, + "pr: #376", + "phase: claimed", + f"head_before: {HEAD_A}", + f"expires_at: {past_expires}", + "profile: prgs-author", + ]) + comments = [{"body": claim_body}] + lease = find_active_conflict_fix_lease(comments, pr_number=376, now=NOW) + self.assertIsNone(lease) + + def test_mismatched_release_different_head(self): + claim_body = _conflict_fix_body(phase="claimed", worktree="branches/fix-376") + expires = (NOW + timedelta(minutes=60)).isoformat().replace("+00:00", "Z") + release_body = "\n".join([ + CONFLICT_FIX_LEASE_MARKER, + "pr: #376", + "profile: prgs-author", + "phase: released", + f"head_before: {HEAD_B}", + f"expires_at: {expires}", + ]) + comments = [{"body": claim_body}, {"body": release_body}] + lease = find_active_conflict_fix_lease(comments, pr_number=376, now=NOW) + self.assertIsNotNone(lease) + self.assertEqual(lease["phase"], "claimed") + + def test_mismatched_release_different_branch(self): + claim_body = "\n".join([ + CONFLICT_FIX_LEASE_MARKER, + "pr: #376", + "branch: feat/branch-A", + "phase: claimed", + f"head_before: {HEAD_A}", + f"expires_at: {(NOW + timedelta(minutes=60)).isoformat().replace('+00:00', 'Z')}", + "profile: prgs-author", + ]) + release_body = "\n".join([ + CONFLICT_FIX_LEASE_MARKER, + "pr: #376", + "branch: feat/branch-B", + "phase: released", + f"head_before: {HEAD_A}", + f"expires_at: {(NOW + timedelta(minutes=60)).isoformat().replace('+00:00', 'Z')}", + "profile: prgs-author", + ]) + comments = [{"body": claim_body}, {"body": release_body}] + lease = find_active_conflict_fix_lease(comments, pr_number=376, now=NOW) + self.assertIsNotNone(lease) + self.assertEqual(lease["phase"], "claimed") + + def test_release_followed_by_newer_claim(self): + claim_1 = _conflict_fix_body(phase="claimed", worktree="branches/fix-376") + expires = (NOW + timedelta(minutes=60)).isoformat().replace("+00:00", "Z") + release_1 = "\n".join([ + CONFLICT_FIX_LEASE_MARKER, + "pr: #376", + "profile: prgs-author", + "phase: released", + f"head_before: {HEAD_A}", + f"head_after: {HEAD_B}", + f"expires_at: {expires}", + ]) + claim_2 = "\n".join([ + CONFLICT_FIX_LEASE_MARKER, + "pr: #376", + "profile: prgs-author", + "phase: claimed", + f"head_before: {HEAD_B}", + f"expires_at: {expires}", + ]) + comments = [{"body": claim_1}, {"body": release_1}, {"body": claim_2}] + lease = find_active_conflict_fix_lease(comments, pr_number=376, now=NOW) + self.assertIsNotNone(lease) + self.assertEqual(lease["head_before"], HEAD_B) + + def test_malformed_or_ambiguous_markers(self): + malformed_release = "\n".join([ + CONFLICT_FIX_LEASE_MARKER, + "pr: #376", + "phase: released", + # missing head_before and profile + ]) + claim_body = _conflict_fix_body(phase="claimed") + comments = [{"body": claim_body}, {"body": malformed_release}] + lease = find_active_conflict_fix_lease(comments, pr_number=376, now=NOW) + self.assertIsNotNone(lease) + + def test_pr818_historical_sequence(self): + comment_14696 = "\n".join([ + "", + "pr: #818", + "branch: feat/issue-638-webui-app-shell-phase1", + "worktree: /Users/jasonwalker/Development/Gitea-Tools/branches/issue-638-webui-app-shell-phase1", + "profile: prgs-author", + "session_id: unknown", + "phase: claimed", + "head_before: 08061b7b8aebdd099a37d1abf5dafcf38e4fd3fb", + "expires_at: 2026-07-23T07:12:13Z", + "reviewer_active: no", + ]) + comment_14730 = "\n".join([ + "", + "pr: #818", + "branch: feat/issue-638-webui-app-shell-phase1", + "worktree: /Users/jasonwalker/Development/Gitea-Tools/branches/issue-638-webui-app-shell-phase1", + "profile: prgs-author", + "session_id: prgs-author-61241-e5129c60", + "phase: released", + "head_before: 08061b7b8aebdd099a37d1abf5dafcf38e4fd3fb", + "head_after: 64b6eb5d5402663098de5ded3b0617cc3b3df98f", + "expires_at: 2026-07-23T06:05:00Z", + "reviewer_active: no", + ]) + comments = [{"body": comment_14696}, {"body": comment_14730}] + check_now = datetime(2026, 7, 23, 6, 30, tzinfo=timezone.utc) + lease = find_active_conflict_fix_lease(comments, pr_number=818, now=check_now) + self.assertIsNone(lease) + + reviewer_gate = assess_reviewer_mutation_blocked( + pr_number=818, + comments=comments, + reviewed_head_sha="64b6eb5d5402663098de5ded3b0617cc3b3df98f", + live_head_sha="64b6eb5d5402663098de5ded3b0617cc3b3df98f", + mutation="approve", + now=check_now, + ) + self.assertTrue(reviewer_gate["mutation_allowed"]) + + if __name__ == "__main__": unittest.main() \ No newline at end of file diff --git a/tests/test_restart_coordinator.py b/tests/test_restart_coordinator.py new file mode 100644 index 0000000..558aaa0 --- /dev/null +++ b/tests/test_restart_coordinator.py @@ -0,0 +1,340 @@ +"""Tests for the MCP restart coordinator and impact analysis (#658). + +Multi-session fixtures exercise every verdict branch: safe, unsafe (live work), +override, and the fail-closed deny on incomplete inventory. Also covers the +critical-section deny path and the new ``ControlPlaneDB.list_sessions``. +""" + +from __future__ import annotations + +import os +import tempfile +import unittest +from datetime import datetime, timedelta, timezone + +import restart_coordinator as rc +from control_plane_db import ControlPlaneDB + + +NOW = datetime(2026, 7, 24, 6, 0, 0, tzinfo=timezone.utc) + + +def _ts(dt: datetime) -> str: + return dt.isoformat() + + +def _live_pid() -> int: + return os.getpid() + + +def _dead_pid() -> int: + # A pid that is essentially never alive. os.kill(0) on it raises + # ProcessLookupError → is_process_alive False. + return 2_000_000_000 + + +def _session(session_id, *, pid, status="active", heartbeat=None, role="author"): + return { + "session_id": session_id, + "role": role, + "profile": "prgs-author", + "pid": pid, + "status": status, + "last_heartbeat_at": _ts(heartbeat or NOW), + } + + +def _lease( + lease_id, + *, + session_id, + freshness, + kind="issue", + number=658, + phase="allocated", + worktree=None, + role="author", +): + return { + "lease_id": lease_id, + "session_id": session_id, + "role": role, + "phase": phase, + "work_kind": kind, + "work_number": number, + "worktree_path": worktree, + "freshness": {"freshness": freshness}, + } + + +class EvaluateRestartImpactTest(unittest.TestCase): + def test_incomplete_inventory_denies_fail_closed(self) -> None: + report = rc.evaluate_restart_impact( + {"inventory_complete": False, "incomplete_reasons": ["db down"]}, + now=NOW, + ) + self.assertEqual(report.verdict, rc.VERDICT_UNSAFE) + self.assertFalse(report.allow_restart) + self.assertFalse(report.restart_performed) + self.assertIn("db down", report.incomplete_reasons) + self.assertTrue( + any("fail closed" in reasoning for reasoning in report.reasons) + ) + + def test_missing_completeness_flag_denies(self) -> None: + # No inventory_complete key at all → treated as incomplete. + report = rc.evaluate_restart_impact({}, now=NOW) + self.assertEqual(report.verdict, rc.VERDICT_UNSAFE) + self.assertFalse(report.allow_restart) + + def test_no_other_work_is_safe(self) -> None: + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [_session("requester", pid=_live_pid())], + "leases": [], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.verdict, rc.VERDICT_SAFE) + self.assertTrue(report.allow_restart) + self.assertEqual(report.blast_radius, rc.BLAST_NONE) + self.assertEqual(report.affected_issues, []) + + def test_dead_foreign_session_and_lease_are_not_disruptive(self) -> None: + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [ + _session("requester", pid=_live_pid()), + _session("dead", pid=_dead_pid()), + ], + "leases": [ + _lease("l-dead", session_id="dead", freshness="stale_dead_process") + ], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.verdict, rc.VERDICT_SAFE) + self.assertTrue(report.allow_restart) + self.assertEqual(report.counts["leases_disruptive"], 0) + self.assertEqual(report.counts["sessions_live_other"], 0) + + def test_live_foreign_lease_denies_without_override(self) -> None: + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [ + _session("requester", pid=_live_pid()), + _session("worker", pid=_live_pid()), + ], + "leases": [ + _lease( + "l1", + session_id="worker", + freshness="active", + worktree="/tmp/wt-658", + phase="implementing", + ) + ], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.verdict, rc.VERDICT_UNSAFE) + self.assertFalse(report.allow_restart) + # Critical section detected: active lease with a live owner. + self.assertEqual(len(report.critical_sections), 1) + self.assertEqual(report.affected_issues, [658]) + self.assertEqual(report.counts["mutations"], 1) + self.assertTrue(report.override_would_allow) + self.assertEqual(report.blast_radius, rc.BLAST_HIGH) + # Placeholder ack state for the affected session. + self.assertEqual(report.ack_state.get("worker"), "pending") + + def test_operator_override_allows_despite_live_work(self) -> None: + inv = { + "inventory_complete": True, + "sessions": [ + _session("requester", pid=_live_pid()), + _session("worker", pid=_live_pid()), + ], + "leases": [_lease("l1", session_id="worker", freshness="active")], + } + report = rc.evaluate_restart_impact( + inv, + now=NOW, + requesting_session_id="requester", + operator_override=True, + ) + self.assertEqual(report.verdict, rc.VERDICT_OVERRIDE) + self.assertTrue(report.allow_restart) + self.assertFalse(report.restart_performed) + + def test_deny_when_critical_section_open(self) -> None: + # A single live author lease in a mutating phase is a critical section + # that must deny an un-overridden restart. + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [_session("worker", pid=_live_pid())], + "leases": [ + _lease( + "l1", + session_id="worker", + freshness="active", + phase="merging", + kind="pr", + number=900, + ) + ], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.verdict, rc.VERDICT_UNSAFE) + self.assertFalse(report.allow_restart) + self.assertEqual(report.affected_prs, [900]) + self.assertEqual(len(report.critical_sections), 1) + + def test_terminal_lock_makes_restart_unsafe(self) -> None: + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [_session("requester", pid=_live_pid())], + "leases": [], + "terminal_lock": {"terminal_pr": 812}, + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.verdict, rc.VERDICT_UNSAFE) + self.assertFalse(report.allow_restart) + self.assertIsNotNone(report.terminal_lock) + self.assertTrue( + any("terminal" in reasoning for reasoning in report.reasons) + ) + + def test_other_live_session_without_lease_is_disruptive(self) -> None: + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [ + _session("requester", pid=_live_pid()), + _session("idle-but-live", pid=_live_pid()), + ], + "leases": [], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.verdict, rc.VERDICT_UNSAFE) + self.assertEqual(report.counts["sessions_live_other"], 1) + + def test_stale_heartbeat_session_not_counted_live(self) -> None: + stale = NOW - timedelta(hours=2) + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [ + _session("requester", pid=_live_pid()), + _session("stale", pid=_live_pid(), heartbeat=stale), + ], + "leases": [], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.verdict, rc.VERDICT_SAFE) + self.assertEqual(report.counts["sessions_live_other"], 0) + + def test_prior_recovery_attempts_echoed(self) -> None: + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [_session("requester", pid=_live_pid())], + "leases": [], + "prior_recovery_attempts": [ + {"kind": "client_reconnect", "at": _ts(NOW)} + ], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(len(report.prior_recovery_attempts), 1) + self.assertEqual(report.counts["prior_recovery_attempts"], 1) + + def test_bare_string_freshness_accepted(self) -> None: + lease = _lease("l1", session_id="worker", freshness="active") + lease["freshness"] = "active" # bare string, not a dict + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [_session("worker", pid=_live_pid())], + "leases": [lease], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.counts["leases_disruptive"], 1) + + def test_as_dict_is_serializable_dto(self) -> None: + import json + + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": [_session("requester", pid=_live_pid())], + "leases": [], + }, + now=NOW, + requesting_session_id="requester", + ) + payload = report.as_dict() + # Round-trips through JSON — safe for the console DTO. + encoded = json.dumps(payload) + decoded = json.loads(encoded) + self.assertEqual(decoded["verdict"], rc.VERDICT_SAFE) + self.assertIn("audit_record", decoded) + self.assertEqual(decoded["audit_record"]["event"], "restart_impact_evaluated") + self.assertFalse(decoded["restart_performed"]) + self.assertIn("coordinator_version", decoded) + + +class ListSessionsTest(unittest.TestCase): + def setUp(self) -> None: + self._tmp = tempfile.TemporaryDirectory() + self.db = ControlPlaneDB(os.path.join(self._tmp.name, "cp.sqlite3")) + + def tearDown(self) -> None: + self._tmp.cleanup() + + def test_list_sessions_filters_by_status(self) -> None: + self.db.upsert_session(session_id="a", role="author", pid=1, status="active") + self.db.upsert_session(session_id="b", role="author", pid=2, status="ended") + active = self.db.list_sessions(statuses=("active",)) + ids = {row["session_id"] for row in active} + self.assertEqual(ids, {"a"}) + every = self.db.list_sessions() + self.assertEqual({row["session_id"] for row in every}, {"a", "b"}) + + def test_list_sessions_feeds_coordinator(self) -> None: + self.db.upsert_session( + session_id="requester", role="author", pid=os.getpid(), status="active" + ) + report = rc.evaluate_restart_impact( + { + "inventory_complete": True, + "sessions": self.db.list_sessions(statuses=("active",)), + "leases": [], + }, + now=NOW, + requesting_session_id="requester", + ) + self.assertEqual(report.counts["sessions_total"], 1) + + +if __name__ == "__main__": # pragma: no cover + unittest.main() diff --git a/tests/test_webui_shell.py b/tests/test_webui_shell.py new file mode 100644 index 0000000..e3c117f --- /dev/null +++ b/tests/test_webui_shell.py @@ -0,0 +1,135 @@ +"""Tests for the Phase 1 operator console application shell (#638).""" +import sys +import unittest +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parent.parent)) + +from starlette.routing import Route +from starlette.testclient import TestClient + +from webui import layout +from webui.app import create_app +from webui.nav import NAV_GROUPS, STUB_PAGES, nav_hrefs + + +class TestShellNav(unittest.TestCase): + def setUp(self): + self.client = TestClient(create_app()) + + def test_nav_group_labels_present(self): + text = self.client.get("/").text + for group in NAV_GROUPS: + with self.subTest(group=group.label): + self.assertIn(f">{group.label}<", text) + + def test_phase1_group_labels_cover_expected_ia(self): + labels = {group.label for group in NAV_GROUPS} + for expected in ( + "Health", + "Traffic", + "Runtime/Sessions", + "Projects", + "Inventory", + "Timeline", + "Policy", + "Insights", + ): + with self.subTest(label=expected): + self.assertIn(expected, labels) + + def test_every_nav_href_resolves_to_a_get_route(self): + app = create_app() + get_paths = { + route.path + for route in app.routes + if isinstance(route, Route) and "GET" in route.methods + } + for href in nav_hrefs(): + with self.subTest(href=href): + self.assertIn(href, get_paths, f"nav href {href} has no GET route") + + def test_legacy_hrefs_still_navigable(self): + text = self.client.get("/").text + for href in ("/queue", "/projects", "/prompts", "/runtime", + "/audit", "/worktrees", "/leases", "/actions"): + with self.subTest(href=href): + self.assertIn(f'href="{href}"', text) + + +class TestShellBadges(unittest.TestCase): + def setUp(self): + self.client = TestClient(create_app()) + + def test_mode_badge_present(self): + self.assertIn("mode: read-only", self.client.get("/").text) + + def test_environment_badge_present(self): + self.assertIn("env:", self.client.get("/").text) + + def test_default_environment_is_local(self): + self.assertEqual(layout.environment_label(), "local") + + def test_remote_bind_reports_remote_environment(self): + import os + + prior = os.environ.get("WEBUI_HOST") + os.environ["WEBUI_HOST"] = "10.0.0.5" + try: + self.assertEqual(layout.environment_label(), "remote") + finally: + if prior is None: + os.environ.pop("WEBUI_HOST", None) + else: + os.environ["WEBUI_HOST"] = prior + + def test_docs_link_present(self): + text = self.client.get("/").text + self.assertIn(layout.DOCS_URL, text) + self.assertIn(">Docs<", text) + + +class TestShellStubs(unittest.TestCase): + def setUp(self): + self.client = TestClient(create_app()) + + def test_stub_routes_render_200(self): + for path, (title, _desc) in STUB_PAGES.items(): + with self.subTest(path=path): + response = self.client.get(path) + self.assertEqual(response.status_code, 200, path) + self.assertIn(title, response.text) + self.assertIn("placeholder", response.text) + + def test_stub_routes_are_read_only(self): + for path in STUB_PAGES: + with self.subTest(path=path): + response = self.client.post(path) + self.assertEqual(response.status_code, 405) + self.assertEqual(response.json()["error"], "read-only-mvp") + + def test_stub_pages_carry_nav_and_badges(self): + response = self.client.get("/inventory") + self.assertIn("mode: read-only", response.text) + self.assertIn('href="/queue"', response.text) + + +class TestShellHome(unittest.TestCase): + def setUp(self): + self.client = TestClient(create_app()) + + def test_home_summarizes_console(self): + text = self.client.get("/").text + self.assertIn("Operator console", text) + self.assertIn("Phase 1", text) + + def test_home_links_legacy_pages(self): + text = self.client.get("/").text + self.assertIn("MVP legacy pages", text) + for href in ("/queue", "/audit", "/leases"): + with self.subTest(href=href): + self.assertIn(f'href="{href}"', text) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/test_webui_system_health_dashboard.py b/tests/test_webui_system_health_dashboard.py new file mode 100644 index 0000000..417e4ae --- /dev/null +++ b/tests/test_webui_system_health_dashboard.py @@ -0,0 +1,345 @@ +"""Tests for the system-health dashboard view (#639). + +Covers the acceptance criteria directly: the page renders the health DTO +fields (AC1), degraded dependencies are visible (AC2), stale runtime is warned +prominently and never rendered as mutation-safe (AC3), healthy and degraded +fixtures both render (AC4), and the shell carries a nav entry (AC5). +""" +import sys +import unittest +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parent.parent)) + +from starlette.testclient import TestClient + +from webui.app import create_app +from webui.deployment_boundary import scan_text_for_client_secrets +from webui.layout import render_page +from webui.nav import iter_nav_items +from webui.system_health import ( + STATUS_DEGRADED, + STATUS_DOWN, + STATUS_OK, + STATUS_SKIPPED, + STATUS_UNPROVEN, + DependencyProbe, + StaleRuntime, + SystemHealthSnapshot, + VersionInfo, +) +from webui.system_health_views import render_system_health_page + +DASHBOARD_PATH = "/system-health" + + +def _version(*, known: bool = True) -> VersionInfo: + return VersionInfo( + git_sha="1c455b6ec0f9cb761fe6248de68c17e061fb5ecd" if known else None, + git_describe="v0.4.1-12-g1c455b6" if known else None, + control_plane_schema_version=4 if known else None, + python_version="3.13.1", + known=known, + ) + + +def _parity(*, stale: bool = False, determinable: bool = True) -> StaleRuntime: + if stale: + return StaleRuntime( + daemon_head="aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa", + checkout_head="bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb", + remote_head="cccccccccccccccccccccccccccccccccccccccc", + stale=True, + determinable=True, + mutation_safe=False, + reasons=("runtime, checkout, and remote commits disagree",), + ) + if not determinable: + return StaleRuntime( + daemon_head=None, + checkout_head=None, + remote_head=None, + stale=False, + determinable=False, + mutation_safe=False, + reasons=("local checkout HEAD could not be read",), + ) + return StaleRuntime( + daemon_head="1c455b6ec0f9cb761fe6248de68c17e061fb5ecd", + checkout_head="1c455b6ec0f9cb761fe6248de68c17e061fb5ecd", + remote_head="1c455b6ec0f9cb761fe6248de68c17e061fb5ecd", + stale=False, + determinable=True, + mutation_safe=True, + reasons=(), + ) + + +def _snapshot( + *, + status: str = STATUS_OK, + ready: bool = True, + readiness_complete: bool = True, + readiness_reasons: tuple[str, ...] = (), + dependencies: tuple[DependencyProbe, ...] | None = None, + parity: StaleRuntime | None = None, + namespaces: tuple[dict, ...] = (), + probe_errors: tuple[str, ...] = (), + version_known: bool = True, +) -> SystemHealthSnapshot: + if dependencies is None: + dependencies = ( + DependencyProbe( + name="control_plane_db", + kind="sqlite", + status=STATUS_OK, + detail="schema version 4", + required=True, + latency_ms=1.25, + metadata={"schema_version": 4}, + ), + ) + return SystemHealthSnapshot( + status=status, + ready=ready, + readiness_complete=readiness_complete, + readiness_reasons=readiness_reasons, + service="mcp-control-plane-webui", + mode="read-only", + version=_version(known=version_known), + started_at="2026-07-23T19:50:47+00:00", + uptime_seconds=3661.5, + timestamp="2026-07-23T20:51:48+00:00", + deep_probes_requested=False, + dependencies=dependencies, + mcp_namespaces=namespaces, + stale_runtime=parity if parity is not None else _parity(), + probe_errors=probe_errors, + ) + + +class TestHealthyRender(unittest.TestCase): + """AC1 / AC4 — every health DTO field reaches the page.""" + + def setUp(self): + self.html = render_system_health_page(_snapshot()) + + def test_readiness_fields_render(self): + self.assertIn("System health", self.html) + self.assertIn("Ready", self.html) + self.assertIn("mcp-control-plane-webui", self.html) + self.assertIn("read-only", self.html) + self.assertIn("2026-07-23T20:51:48+00:00", self.html) + + def test_version_and_uptime_render(self): + self.assertIn("1c455b6ec0f9cb761fe6248de68c17e061fb5ecd", self.html) + self.assertIn("v0.4.1-12-g1c455b6", self.html) + self.assertIn("3.13.1", self.html) + self.assertIn("3661.500s", self.html) + self.assertIn("1.02h", self.html) + + def test_dependency_row_renders_with_latency(self): + self.assertIn("control_plane_db", self.html) + self.assertIn("sqlite", self.html) + self.assertIn("schema version 4", self.html) + self.assertIn("1.2 ms", self.html) + + def test_healthy_page_shows_no_stale_warning(self): + self.assertNotIn("Stale runtime:", self.html) + self.assertNotIn("Staleness", self.html) + + def test_unknown_version_is_labelled_not_faked(self): + html = render_system_health_page(_snapshot(version_known=False)) + self.assertIn("unknown", html) + self.assertIn("unresolved", html) + + +class TestDegradedRender(unittest.TestCase): + """AC2 — a degraded or unrun dependency is visible, not swallowed.""" + + def setUp(self): + self.deps = ( + DependencyProbe( + name="control_plane_db", + kind="sqlite", + status=STATUS_OK, + detail="schema version 4", + required=True, + latency_ms=0.9, + ), + DependencyProbe( + name="repository", + kind="git", + status=STATUS_DOWN, + detail="repository root is not a git checkout", + required=True, + latency_ms=4.0, + ), + DependencyProbe( + name="gitea", + kind="http", + status=STATUS_SKIPPED, + detail="deep probe not requested", + required=False, + ), + ) + self.html = render_system_health_page( + _snapshot( + status=STATUS_DEGRADED, + ready=False, + readiness_complete=False, + readiness_reasons=("required dependency 'repository' is down",), + dependencies=self.deps, + ) + ) + + def test_degraded_banner_names_the_dependency(self): + self.assertIn("Degraded dependencies:", self.html) + self.assertIn("repository", self.html) + + def test_not_run_probe_is_reported_separately(self): + self.assertIn("Not probed:", self.html) + self.assertIn("gitea", self.html) + self.assertIn("not counted", self.html) + + def test_not_ready_headline_and_reason(self): + self.assertIn("Not ready", self.html) + self.assertIn("required dependency 'repository' is down", self.html) + + def test_degraded_status_badge_present(self): + self.assertIn("badge-health-degraded", self.html) + self.assertIn("badge-health-down", self.html) + + def test_ready_but_incomplete_is_not_shown_as_plain_ready(self): + html = render_system_health_page( + _snapshot(ready=True, readiness_complete=False) + ) + self.assertIn("Ready (incomplete evidence)", html) + + +class TestStaleRuntimeWarning(unittest.TestCase): + """AC3 — staleness is prominent and never claims mutation safety.""" + + def test_stale_runtime_warns_and_denies_mutation_safety(self): + html = render_system_health_page(_snapshot(parity=_parity(stale=True))) + self.assertIn("Stale runtime:", html) + self.assertIn("do not treat this runtime as mutation-safe", html) + self.assertIn("Mutation safeFalse", html) + + def test_indeterminate_parity_is_not_reported_safe(self): + html = render_system_health_page( + _snapshot(parity=_parity(determinable=False)) + ) + self.assertIn("Staleness", html) + self.assertIn("Mutation safeFalse", html) + self.assertIn("DeterminableFalse", html) + + def test_healthy_parity_reports_mutation_safe_true(self): + html = render_system_health_page(_snapshot()) + self.assertIn("Mutation safeTrue", html) + + +class TestNamespacesAndErrors(unittest.TestCase): + def test_unproven_namespace_rows_render(self): + html = render_system_health_page( + _snapshot( + namespaces=( + { + "namespace": "gitea-author", + "required_tool": "gitea_lock_issue", + "status": STATUS_UNPROVEN, + "ide_namespace_proven": False, + "reason": "the web console cannot invoke the IDE-managed MCP client", + }, + ) + ) + ) + self.assertIn("gitea-author", html) + self.assertIn("gitea_lock_issue", html) + self.assertIn("badge-health-unproven", html) + + def test_no_namespaces_degrades_gracefully(self): + html = render_system_health_page(_snapshot(namespaces=())) + self.assertIn("No MCP namespaces are declared.", html) + + def test_probe_errors_render_when_present(self): + html = render_system_health_page( + _snapshot(probe_errors=("probe raised: disk offline",)) + ) + self.assertIn("Probe errors", html) + self.assertIn("disk offline", html) + + def test_probe_error_card_absent_when_clean(self): + self.assertNotIn("Probe errors", render_system_health_page(_snapshot())) + + +class TestReadOnlyAndRedaction(unittest.TestCase): + def test_no_restart_or_kill_controls(self): + html = render_system_health_page(_snapshot()) + self.assertNotIn("", html) + self.assertIn("<script>", html) + + +class TestNavAndRoute(unittest.TestCase): + """AC5 — the shell links the dashboard, and the route serves it.""" + + def setUp(self): + self.client = TestClient(create_app()) + + def test_nav_contains_system_health(self): + self.assertIn( + (DASHBOARD_PATH, "System health"), + [(item.href, item.label) for item in iter_nav_items()], + ) + + def test_rendered_shell_links_dashboard(self): + page = render_page(title="Home", body_html="

x

") + self.assertIn(f'href="{DASHBOARD_PATH}"', page) + + def test_route_renders_dashboard(self): + response = self.client.get(DASHBOARD_PATH) + self.assertEqual(response.status_code, 200) + self.assertIn("System health", response.text) + self.assertIn("Stale-runtime parity", response.text) + + def test_route_is_read_only(self): + self.assertEqual(self.client.post(DASHBOARD_PATH).status_code, 405) + + def test_live_page_leaks_no_client_secret(self): + findings = scan_text_for_client_secrets(self.client.get(DASHBOARD_PATH).text) + self.assertEqual(findings, []) + + +if __name__ == "__main__": # pragma: no cover + unittest.main() diff --git a/tests/test_webui_timeline.py b/tests/test_webui_timeline.py new file mode 100644 index 0000000..39c7ea5 --- /dev/null +++ b/tests/test_webui_timeline.py @@ -0,0 +1,1021 @@ +"""Tests for the workflow-event timeline model and read API (#637). + +Covers the acceptance criteria: versioned schema, adaptation of control-plane +events and Gitea handoff comments, filter by issue/PR/session, redaction of +secret-like payloads, and stable pagination. +""" +import json +import os +import sqlite3 +import sys +import unittest +from pathlib import Path + +sys.path.insert(0, str(Path(__file__).resolve().parent.parent)) + +from starlette.testclient import TestClient + +import control_plane_db +from canonical_thread_handoff import ( + CTH_TYPES, + MARKER, + format_cth_body, + is_known_cth_type, + parse_cth_comment, +) +from webui import timeline +from webui.app import create_app + + +def _seed_db(path: str) -> None: + """Create a control-plane DB and seed scoped work_items + events.""" + # Constructing ControlPlaneDB runs the schema migration once. + control_plane_db.ControlPlaneDB(db_path=path) + conn = sqlite3.connect(path) + try: + conn.execute( + "INSERT INTO work_items(remote, org, repo, kind, number, state, updated_at) " + "VALUES (?,?,?,?,?,?,?)", + ("prgs", "Scaled-Tech-Consulting", "Gitea-Tools", "issue", 637, "open", "2026-07-23T00:00:00Z"), + ) + issue_wid = conn.execute("SELECT last_insert_rowid()").fetchone()[0] + conn.execute( + "INSERT INTO work_items(remote, org, repo, kind, number, state, updated_at) " + "VALUES (?,?,?,?,?,?,?)", + ("prgs", "Scaled-Tech-Consulting", "Gitea-Tools", "pr", 813, "open", "2026-07-23T00:00:00Z"), + ) + pr_wid = conn.execute("SELECT last_insert_rowid()").fetchone()[0] + # A work item for a different repo — must never appear in prgs/Gitea-Tools scope. + conn.execute( + "INSERT INTO work_items(remote, org, repo, kind, number, state, updated_at) " + "VALUES (?,?,?,?,?,?,?)", + ("dadeschools", "Other", "Elsewhere", "issue", 1, "open", "2026-07-23T00:00:00Z"), + ) + other_wid = conn.execute("SELECT last_insert_rowid()").fetchone()[0] + + rows = [ + (issue_wid, "allocation", "assigned author work", "2026-07-23T01:00:00Z"), + (issue_wid, "lease.renew", "token=ghs_ABCDEF1234567890abcdef lease renewed", "2026-07-23T02:00:00Z"), + (pr_wid, "pr.opened", "PR opened for review", "2026-07-23T03:00:00Z"), + (other_wid, "allocation", "off-scope event", "2026-07-23T04:00:00Z"), + ] + conn.executemany( + "INSERT INTO events(work_item_id, event_type, message, created_at) VALUES (?,?,?,?)", + rows, + ) + conn.commit() + finally: + conn.close() + + +class TestSchema(unittest.TestCase): + def test_schema_is_versioned(self): + self.assertIsInstance(timeline.TIMELINE_SCHEMA_VERSION, int) + self.assertGreaterEqual(timeline.TIMELINE_SCHEMA_VERSION, 1) + + def test_event_to_dict_shape(self): + ev = timeline.WorkflowEvent( + source=timeline.SOURCE_CONTROL_PLANE, + event_type="allocation", + event_key="cp:1", + timestamp="2026-07-23T01:00:00Z", + issue_number=637, + ) + d = ev.to_dict() + for key in ( + "source", "event_type", "event_key", "timestamp", "actor", "role", + "issue_number", "pr_number", "session_id", "tool_name", "decision", + "message", "correlation_id", "evidence_refs", "sensitive", + ): + self.assertIn(key, d) + self.assertEqual(d["evidence_refs"], []) + + +class TestCpAdapter(unittest.TestCase): + def test_issue_and_pr_mapping(self): + rows = [ + {"event_id": 1, "event_type": "allocation", "message": "x", "created_at": "2026-07-23T01:00:00Z", "kind": "issue", "number": 637}, + {"event_id": 2, "event_type": "pr.opened", "message": "y", "created_at": "2026-07-23T02:00:00Z", "kind": "pr", "number": 813}, + ] + events = timeline.adapt_cp_events(rows) + self.assertEqual(len(events), 2) + self.assertEqual(events[0].issue_number, 637) + self.assertIsNone(events[0].pr_number) + self.assertEqual(events[0].correlation_id, "issue#637") + self.assertIsNone(events[1].issue_number) + self.assertEqual(events[1].pr_number, 813) + + def test_malformed_rows_skipped(self): + rows = [ + {"event_id": None, "event_type": "x", "kind": "issue", "number": 1}, + {"event_id": 5, "event_type": "", "kind": "issue", "number": 1}, + {"event_id": 6, "event_type": "ok", "message": "m", "created_at": None, "kind": "issue", "number": 1}, + ] + events = timeline.adapt_cp_events(rows) + self.assertEqual(len(events), 1) + self.assertIsNone(events[0].timestamp) + + def test_sensitive_event_flagged(self): + rows = [{"event_id": 1, "event_type": "lease.renew", "message": "m", "created_at": "2026-07-23T01:00:00Z", "kind": "issue", "number": 1}] + events = timeline.adapt_cp_events(rows) + self.assertTrue(events[0].sensitive) + + +class TestCthAdapter(unittest.TestCase): + def test_cth_comment_becomes_event(self): + body = format_cth_body( + cth_type="Author Handoff", + status="ready", + next_owner="reviewer", + decision="implement timeline", + proof="commit abc1234 closes #637", + next_action="review PR", + ready_to_paste_prompt="Review PR #900 as reviewer", + ) + comments = [{"id": 42, "body": body, "created_at": "2026-07-23T05:00:00Z", "user": {"login": "jcwalker3"}}] + events = timeline.adapt_cth_comments(comments, kind="issue", number=637) + self.assertEqual(len(events), 1) + ev = events[0] + self.assertEqual(ev.source, timeline.SOURCE_GITEA_HANDOFF) + self.assertEqual(ev.event_type, "handoff:Author Handoff") + self.assertEqual(ev.actor, "jcwalker3") + self.assertEqual(ev.issue_number, 637) + self.assertEqual(ev.event_key, "cth:issue:637:42") + self.assertIn("#637", ev.evidence_refs) + self.assertIn("abc1234", ev.evidence_refs) + + def test_non_cth_comment_ignored(self): + comments = [{"id": 1, "body": "just a normal comment", "created_at": "2026-07-23T05:00:00Z", "user": {"login": "x"}}] + self.assertEqual(timeline.adapt_cth_comments(comments, kind="issue", number=1), []) + + +class TestRedaction(unittest.TestCase): + def test_cp_message_redacted(self): + rows = [{"event_id": 1, "event_type": "lease", "message": "token=ghs_ABCDEF1234567890abcdef here", "created_at": "2026-07-23T01:00:00Z", "kind": "issue", "number": 1}] + events = timeline.adapt_cp_events(rows) + self.assertNotIn("ghs_ABCDEF1234567890abcdef", events[0].message or "") + + def test_handoff_decision_redacted(self): + body = format_cth_body( + cth_type="Blocker", + status="blocked", + next_owner="author", + decision="password=SuperSecret123! must rotate", + proof="none", + next_action="rotate", + ready_to_paste_prompt="Rotate the credential and retry", + ) + comments = [{"id": 7, "body": body, "created_at": "2026-07-23T05:00:00Z", "user": {"login": "x"}}] + events = timeline.adapt_cth_comments(comments, kind="issue", number=1) + self.assertNotIn("SuperSecret123!", events[0].decision or "") + + +class TestFilterSortPaginate(unittest.TestCase): + def _events(self): + return [ + timeline.WorkflowEvent(source="control_plane", event_type="a", event_key="cp:3", timestamp="2026-07-23T03:00:00Z", pr_number=813), + timeline.WorkflowEvent(source="control_plane", event_type="b", event_key="cp:1", timestamp="2026-07-23T01:00:00Z", issue_number=637), + timeline.WorkflowEvent(source="control_plane", event_type="c", event_key="cp:2", timestamp="2026-07-23T02:00:00Z", issue_number=637, session_id="sess-1"), + ] + + def test_filter_by_issue(self): + out = timeline.filter_events(self._events(), issue=637) + self.assertEqual({e.event_key for e in out}, {"cp:1", "cp:2"}) + + def test_filter_by_pr(self): + out = timeline.filter_events(self._events(), pr=813) + self.assertEqual([e.event_key for e in out], ["cp:3"]) + + def test_filter_by_session(self): + out = timeline.filter_events(self._events(), session="sess-1") + self.assertEqual([e.event_key for e in out], ["cp:2"]) + + def test_stable_sort_ascending(self): + out = timeline.sort_events(self._events()) + self.assertEqual([e.event_key for e in out], ["cp:1", "cp:2", "cp:3"]) + + def test_missing_timestamp_sorts_last(self): + evs = self._events() + [ + timeline.WorkflowEvent(source="control_plane", event_type="z", event_key="cp:9", timestamp=None) + ] + out = timeline.sort_events(evs) + self.assertEqual(out[-1].event_key, "cp:9") + + def test_pagination_windows_and_next_offset(self): + evs = timeline.sort_events(self._events()) + page1 = timeline.paginate(evs, limit=2, offset=0) + self.assertEqual(len(page1.events), 2) + self.assertEqual(page1.total, 3) + self.assertEqual(page1.next_offset, 2) + page2 = timeline.paginate(evs, limit=2, offset=2) + self.assertEqual(len(page2.events), 1) + self.assertIsNone(page2.next_offset) + + def test_pagination_bounds_coerced(self): + evs = self._events() + page = timeline.paginate(evs, limit=-5, offset=-3) + self.assertGreaterEqual(page.limit, 1) + self.assertEqual(page.offset, 0) + + +class TestCpReader(unittest.TestCase): + def test_reads_scoped_events_only(self): + import tempfile + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + events, status = timeline.read_cp_events( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", db_path=db + ) + self.assertTrue(status.ok) + # 3 scoped events; the dadeschools/Other event is excluded. + self.assertEqual(len(events), 3) + self.assertTrue(all(e.source == "control_plane" for e in events)) + # Redaction applied to the token-bearing message. + joined = " ".join(e.message or "" for e in events) + self.assertNotIn("ghs_ABCDEF1234567890abcdef", joined) + + def test_missing_db_degrades(self): + events, status = timeline.read_cp_events( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + db_path="/nonexistent/path/to/cp.sqlite3", + ) + self.assertEqual(events, []) + self.assertFalse(status.ok) + self.assertIsNotNone(status.reason) + + +class TestLoadTimeline(unittest.TestCase): + def test_handoff_not_run_without_thread_filter(self): + import tempfile + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", db_path=db + ) + d = snap.to_dict() + handoff = [s for s in d["sources"] if s["name"] == "gitea_handoff"][0] + self.assertFalse(handoff["ok"]) + self.assertIn("thread-scoped", handoff["reason"]) + self.assertEqual(d["schema_version"], timeline.TIMELINE_SCHEMA_VERSION) + + def test_handoff_included_via_injected_source(self): + import tempfile + + body = format_cth_body( + cth_type="Author Handoff", status="ready", next_owner="reviewer", + decision="d", proof="#637", next_action="review", ready_to_paste_prompt="Review PR #1 now", + ) + + def source(kind, number): + return [{"id": 1, "body": body, "created_at": "2026-07-23T09:00:00Z", "user": {"login": "jcwalker3"}}] + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, db_path=db, comment_source=source, + ) + d = snap.to_dict() + handoff = [s for s in d["sources"] if s["name"] == "gitea_handoff"][0] + self.assertTrue(handoff["ok"]) + self.assertEqual(handoff["count"], 1) + # Both a CP event and the handoff event for issue 637 appear, sorted. + kinds = {e["source"] for e in d["events"]} + self.assertEqual(kinds, {"control_plane", "gitea_handoff"}) + + def test_failing_comment_source_degrades_only_handoff(self): + import tempfile + + def boom(kind, number): + raise RuntimeError("network down") + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, db_path=db, comment_source=boom, + ) + d = snap.to_dict() + cp = [s for s in d["sources"] if s["name"] == "control_plane"][0] + handoff = [s for s in d["sources"] if s["name"] == "gitea_handoff"][0] + self.assertTrue(cp["ok"]) + self.assertFalse(handoff["ok"]) + self.assertIn("network down", handoff["reason"]) + + +# A fabricated 40-character lowercase hex value with the shape of a Gitea +# personal access token. Never a real credential — its only job is to prove it +# cannot reach any part of a serialized timeline payload. +SYNTHETIC_SECRET_40_HEX = "a3f9c17be44d2058e6b17c9d0f5321ab77c4e9d1" + + +def _cth_comment(comment_id, *, created_at, session=None, decision="d", proof="none", **kw): + """Build one real CTH comment record, optionally declaring a session.""" + extra = {"Session": session} if session is not None else None + body = format_cth_body( + cth_type=kw.pop("cth_type", "Author Handoff"), + status=kw.pop("status", "ready"), + next_owner=kw.pop("next_owner", "reviewer"), + decision=decision, + proof=proof, + next_action=kw.pop("next_action", "review"), + ready_to_paste_prompt=kw.pop("ready_to_paste_prompt", "Review PR #1 now"), + extra_fields=extra, + ) + return { + "id": comment_id, + "body": body, + "created_at": created_at, + "user": {"login": kw.pop("login", "jcwalker3")}, + } + + +class TestSessionFilterThroughAdapter(unittest.TestCase): + """F1: the session dimension must be real, or explicitly refused. + + These drive the filter through the CTH adapter and the composed + ``load_timeline``/API path, never through a hand-built ``WorkflowEvent``. + """ + + def _source(self, comments): + return lambda kind, number: list(comments) + + def test_adapter_populates_declared_session(self): + events = timeline.adapt_cth_comments( + [_cth_comment(1, created_at="2026-07-23T05:00:00Z", session="sess-alpha")], + kind="issue", + number=637, + ) + self.assertEqual(len(events), 1) + self.assertEqual(events[0].session_id, "sess-alpha") + + def test_session_filter_matches_through_adapter(self): + import tempfile + + comments = [ + _cth_comment(1, created_at="2026-07-23T09:00:00Z", session="sess-alpha"), + _cth_comment(2, created_at="2026-07-23T08:00:00Z", session="sess-alpha"), + _cth_comment(3, created_at="2026-07-23T10:00:00Z", session="sess-beta"), + ] + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, session="sess-alpha", db_path=db, + comment_source=self._source(comments), + ) + d = snap.to_dict() + self.assertTrue(d["ok"]) + self.assertIsNone(d["error"]) + keys = [e["event_key"] for e in d["events"]] + # Only the two sess-alpha events, still in ascending timestamp order. + self.assertEqual(keys, ["cth:issue:637:2", "cth:issue:637:1"]) + self.assertTrue(all(e["session_id"] == "sess-alpha" for e in d["events"])) + self.assertEqual(d["pagination"]["total"], 2) + + def test_session_filter_paginates_and_keeps_order(self): + import tempfile + + comments = [ + _cth_comment(i, created_at=f"2026-07-23T0{i}:00:00Z", session="sess-alpha") + for i in range(1, 4) + ] + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + kwargs = dict( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, session="sess-alpha", db_path=db, + comment_source=self._source(comments), + ) + page1 = timeline.load_timeline(limit=2, offset=0, **kwargs).to_dict() + page2 = timeline.load_timeline(limit=2, offset=2, **kwargs).to_dict() + self.assertEqual( + [e["event_key"] for e in page1["events"]], + ["cth:issue:637:1", "cth:issue:637:2"], + ) + self.assertEqual(page1["pagination"]["next_offset"], 2) + self.assertEqual([e["event_key"] for e in page2["events"]], ["cth:issue:637:3"]) + self.assertIsNone(page2["pagination"]["next_offset"]) + + def test_unknown_session_is_honestly_empty_when_supported(self): + """A source that *can* answer the dimension may legitimately match nothing.""" + import tempfile + + comments = [_cth_comment(1, created_at="2026-07-23T09:00:00Z", session="sess-alpha")] + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + d = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, session="sess-nope", db_path=db, + comment_source=self._source(comments), + ).to_dict() + self.assertTrue(d["ok"]) + self.assertEqual(d["events"], []) + + def test_session_filter_refused_when_no_source_can_answer(self): + """The F1 defect: an empty-and-healthy page for an unanswerable filter.""" + import tempfile + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + d = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, session="sess-alpha", db_path=db, comment_source=None, + ).to_dict() + self.assertFalse(d["ok"]) + self.assertEqual(d["error"]["code"], "filter_not_supported") + self.assertEqual(d["error"]["unsupported_filters"], ["session"]) + self.assertEqual(d["events"], []) + self.assertEqual(d["pagination"]["total"], 0) + # The refusal states which source could not answer, and why. + explained = {r["source"] for r in d["error"]["sources"]} + self.assertEqual(explained, {"control_plane", "gitea_handoff"}) + + def test_control_plane_declares_session_unsupported(self): + import tempfile + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + d = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, session="sess-alpha", db_path=db, + comment_source=self._source([]), + ).to_dict() + cp = [s for s in d["sources"] if s["name"] == "control_plane"][0] + handoff = [s for s in d["sources"] if s["name"] == "gitea_handoff"][0] + self.assertNotIn("session", cp["supported_filters"]) + self.assertEqual(cp["unsupported_filters"], ["session"]) + self.assertIn("session", handoff["supported_filters"]) + self.assertEqual(handoff["unsupported_filters"], []) + + def test_secret_shaped_session_value_is_dropped(self): + events = timeline.adapt_cth_comments( + [_cth_comment(1, created_at="2026-07-23T05:00:00Z", session=SYNTHETIC_SECRET_40_HEX)], + kind="issue", + number=637, + ) + self.assertIsNone(events[0].session_id) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(events[0].to_dict())) + + +class TestEvidenceRefRedaction(unittest.TestCase): + """F2: evidence_refs must not be a hole in the redaction boundary.""" + + def _refs_for(self, *, proof="none", decision="d"): + events = timeline.adapt_cth_comments( + [_cth_comment(1, created_at="2026-07-23T05:00:00Z", proof=proof, decision=decision)], + kind="issue", + number=637, + ) + self.assertEqual(len(events), 1) + return events[0] + + def test_assigned_secret_never_reaches_evidence_refs(self): + ev = self._refs_for(proof=f"authenticated with token={SYNTHETIC_SECRET_40_HEX}") + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, ev.evidence_refs) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(ev.to_dict())) + + def test_bare_secret_shaped_value_never_reaches_evidence_refs(self): + # An undeclared hex run in proof text is not evidence of anything, and + # proof itself is never serialized — so the value has no way out. + ev = self._refs_for(proof=f"proof {SYNTHETIC_SECRET_40_HEX}", decision="rotate the credential") + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, ev.evidence_refs) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(ev.to_dict())) + + def test_secret_absent_from_complete_serialized_payload(self): + import tempfile + + comments = [ + _cth_comment( + 1, + created_at="2026-07-23T09:00:00Z", + proof=f"lease token={SYNTHETIC_SECRET_40_HEX} and bare {SYNTHETIC_SECRET_40_HEX}", + decision=f"rotate api_key={SYNTHETIC_SECRET_40_HEX}", + ) + ] + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, db_path=db, comment_source=lambda k, n: comments, + ) + payload = json.dumps(snap.to_dict()) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, payload) + # And the surface is clean by the redaction policy's own detectors. + from webui import console_redaction + + self.assertEqual(console_redaction.scan_for_secrets(snap.to_dict()), []) + + def test_legitimate_references_still_usable(self): + head = "4f3a464a1c455b6ecaaae9b6eef496c8f8ed451a" + ev = self._refs_for( + proof=f"closes #637, PR #849, commit abc1234, at head {head}", + decision="none", + ) + for token in ("#637", "#849", "abc1234", head): + self.assertIn(token, ev.evidence_refs) + + def test_undeclared_hex_words_are_not_references(self): + ev = self._refs_for(proof="the record was defaced and the facade decayed") + self.assertEqual(ev.evidence_refs, ()) + + def test_refs_revalidated_independently_before_serialization(self): + """Extraction is not trusted: the validator drops anything unproven.""" + safe, dropped = timeline._validated_evidence_refs( + ["#637", "abc1234", "not-a-ref", f"token={SYNTHETIC_SECRET_40_HEX}", ""] + ) + self.assertEqual(safe, ("#637", "abc1234")) + self.assertTrue(dropped) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(list(safe))) + + def test_clean_event_is_not_marked_sensitive(self): + ev = self._refs_for(proof="closes #637") + self.assertEqual(ev.evidence_refs, ("#637",)) + self.assertFalse(ev.sensitive) + + +class TestTimelineApi(unittest.TestCase): + def setUp(self): + self._prev_db = os.environ.get(control_plane_db.DB_PATH_ENV) + self._prev_offline = os.environ.get("WEBUI_TEST_OFFLINE") + import tempfile + + self._tmpdir = tempfile.TemporaryDirectory() + self._db = os.path.join(self._tmpdir.name, "cp.sqlite3") + _seed_db(self._db) + os.environ[control_plane_db.DB_PATH_ENV] = self._db + os.environ["WEBUI_TEST_OFFLINE"] = "1" + self.client = TestClient(create_app()) + + def tearDown(self): + if self._prev_db is None: + os.environ.pop(control_plane_db.DB_PATH_ENV, None) + else: + os.environ[control_plane_db.DB_PATH_ENV] = self._prev_db + if self._prev_offline is None: + os.environ.pop("WEBUI_TEST_OFFLINE", None) + else: + os.environ["WEBUI_TEST_OFFLINE"] = self._prev_offline + self._tmpdir.cleanup() + + def test_api_returns_timeline(self): + resp = self.client.get("/api/v1/timeline") + self.assertEqual(resp.status_code, 200) + body = resp.json() + self.assertEqual(body["schema_version"], timeline.TIMELINE_SCHEMA_VERSION) + self.assertIn("events", body) + self.assertIn("pagination", body) + self.assertGreaterEqual(body["pagination"]["total"], 1) + + def test_api_filter_by_issue(self): + resp = self.client.get("/api/v1/timeline?issue=637") + self.assertEqual(resp.status_code, 200) + events = resp.json()["events"] + self.assertTrue(events) + self.assertTrue(all(e["issue_number"] == 637 for e in events)) + + def test_api_pagination(self): + resp = self.client.get("/api/v1/timeline?limit=1&offset=0") + self.assertEqual(resp.status_code, 200) + pg = resp.json()["pagination"] + self.assertEqual(pg["limit"], 1) + self.assertEqual(len(resp.json()["events"]), 1) + if pg["total"] > 1: + self.assertTrue(pg["has_more"]) + + def test_api_is_read_only(self): + resp = self.client.post("/api/v1/timeline") + self.assertIn(resp.status_code, (404, 405)) + + def test_api_no_secret_leak(self): + resp = self.client.get("/api/v1/timeline?issue=637") + self.assertNotIn("ghs_ABCDEF1234567890abcdef", resp.text) + + def test_api_refuses_unanswerable_session_filter(self): + """No handoff source is configured offline, so nothing can carry a session.""" + resp = self.client.get("/api/v1/timeline?issue=637&session=sess-alpha") + self.assertEqual(resp.status_code, 422) + body = resp.json() + self.assertFalse(body["ok"]) + self.assertEqual(body["error"]["code"], "filter_not_supported") + self.assertEqual(body["error"]["unsupported_filters"], ["session"]) + self.assertEqual(body["events"], []) + self.assertEqual(body["pagination"]["total"], 0) + + def test_api_unfiltered_read_stays_ok(self): + body = self.client.get("/api/v1/timeline?issue=637").json() + self.assertTrue(body["ok"]) + self.assertIsNone(body["error"]) + + +def _seed_event(path: str, *, event_type: str, message: str, created_at: str) -> None: + """Append one control-plane event with a caller-chosen ``event_type``. + + The ``events`` table stores whatever a producer writes, so this seeds the + adapter the way a hostile or buggy producer would. + """ + conn = sqlite3.connect(path) + try: + wid = conn.execute( + "SELECT work_item_id FROM work_items WHERE kind='issue' AND number=637" + ).fetchone()[0] + conn.execute( + "INSERT INTO events(work_item_id, event_type, message, created_at) VALUES (?,?,?,?)", + (wid, event_type, message, created_at), + ) + conn.commit() + finally: + conn.close() + + +def _raw_cth_body(heading: str, **fields) -> str: + """Build a CTH comment with an arbitrary heading. + + ``format_cth_body`` refuses an undeclared type, which is exactly the write + path already covered. The read path must cope with a body that never went + through it, so this writes the marker and heading directly. + """ + lines = [MARKER, f"## CTH: {heading}", ""] + base = { + "Status": "ready", + "Next owner": "reviewer", + "Current blocker": "none", + "Decision": "d", + "Proof": "none", + "Next action": "review", + "Ready-to-paste prompt": "Review PR #1 now", + } + base.update(fields) + lines.extend(f"{key}: {value}" for key, value in base.items()) + return "\n".join(lines) + + +class TestEventTypeBoundary(unittest.TestCase): + """F2 residual: event_type must cross the same boundary as every other field. + + The canary is seeded through ``event_type`` *only*, with a benign message, + so message redaction cannot be what makes these pass. Each case inspects + ``event_type`` explicitly and then the complete serialized payload. + """ + + # ---- control-plane stored event_type ---------------------------------- # + + def test_cp_event_type_canary_never_serialized(self): + rows = [{ + "event_id": 1, + "event_type": SYNTHETIC_SECRET_40_HEX, + "message": "deploy completed", # benign: no redaction happens here + "created_at": "2026-07-23T01:00:00Z", + "kind": "issue", + "number": 637, + }] + events = timeline.adapt_cp_events(rows) + self.assertEqual(len(events), 1) + ev = events[0] + # The field itself, inspected directly. + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, ev.event_type) + self.assertEqual(ev.event_type, timeline.UNSAFE_EVENT_TYPE) + # Message redaction is provably not what saved us: it is untouched. + self.assertEqual(ev.message, "deploy completed") + # And nowhere in the serialized record. + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(ev.to_dict())) + # An unsafe value is a visible fact, not a silent substitution. + self.assertTrue(ev.sensitive) + + def test_cp_same_canary_in_message_and_event_type(self): + """The decisive case: one record, one value, two fields, one verdict.""" + rows = [{ + "event_id": 2, + "event_type": SYNTHETIC_SECRET_40_HEX, + "message": f"deploy token={SYNTHETIC_SECRET_40_HEX}", + "created_at": "2026-07-23T01:00:00Z", + "kind": "issue", + "number": 637, + }] + ev = timeline.adapt_cp_events(rows)[0] + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, ev.message or "") + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, ev.event_type) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(ev.to_dict())) + + def test_cp_unsafe_event_type_is_not_rewritten_as_a_valid_one(self): + """A refused value must not be disguised as some other real event type.""" + rows = [{ + "event_id": 3, + "event_type": SYNTHETIC_SECRET_40_HEX, + "message": "m", + "created_at": "2026-07-23T01:00:00Z", + "kind": "issue", + "number": 637, + }] + ev = timeline.adapt_cp_events(rows)[0] + for legitimate in ("allocation", "pr.opened", "lease.renew", "assigned"): + self.assertNotEqual(ev.event_type, legitimate) + + def test_cp_assigned_secret_in_event_type_refused(self): + rows = [{ + "event_id": 4, + "event_type": f"token={SYNTHETIC_SECRET_40_HEX}", + "message": "m", + "created_at": "2026-07-23T01:00:00Z", + "kind": "issue", + "number": 637, + }] + ev = timeline.adapt_cp_events(rows)[0] + self.assertEqual(ev.event_type, timeline.UNSAFE_EVENT_TYPE) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(ev.to_dict())) + + def test_cp_legitimate_event_types_preserved(self): + """Every real producer type in this repo must survive untouched.""" + legitimate = [ + "allocation", "pr.opened", "lease.renew", "assigned", + "lease_released", "lease_expired", "lease_abandoned", + "lease_adopted", "dependency_edge_state_change", + ] + rows = [ + { + "event_id": i, + "event_type": name, + "message": "m", + "created_at": "2026-07-23T01:00:00Z", + "kind": "issue", + "number": 637, + } + for i, name in enumerate(legitimate, start=1) + ] + events = timeline.adapt_cp_events(rows) + self.assertEqual([e.event_type for e in events], legitimate) + + def test_cp_malformed_event_type_shapes_refused(self): + for bad in ("has space", "1leading-digit", "x" * 200, "semi;colon", "new\nline"): + with self.subTest(event_type=bad): + rows = [{ + "event_id": 1, + "event_type": bad, + "message": "m", + "created_at": "2026-07-23T01:00:00Z", + "kind": "issue", + "number": 637, + }] + events = timeline.adapt_cp_events(rows) + self.assertEqual(events[0].event_type, timeline.UNSAFE_EVENT_TYPE) + self.assertNotIn(bad, json.dumps(events[0].to_dict())) + + # ---- CTH heading event_type ------------------------------------------- # + + def test_cth_heading_canary_never_serialized(self): + comments = [{ + "id": 11, + "body": _raw_cth_body(SYNTHETIC_SECRET_40_HEX), + "created_at": "2026-07-23T05:00:00Z", + "user": {"login": "jcwalker3"}, + }] + events = timeline.adapt_cth_comments(comments, kind="issue", number=637) + self.assertEqual(len(events), 1) + ev = events[0] + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, ev.event_type) + self.assertEqual(ev.event_type, timeline.UNKNOWN_HANDOFF_EVENT_TYPE) + # No message redaction is doing the work here — the message is benign. + self.assertEqual(ev.message, "review") + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(ev.to_dict())) + self.assertTrue(ev.sensitive) + + def test_cth_assigned_secret_heading_refused(self): + comments = [{ + "id": 12, + "body": _raw_cth_body(f"token={SYNTHETIC_SECRET_40_HEX}"), + "created_at": "2026-07-23T05:00:00Z", + "user": {"login": "jcwalker3"}, + }] + ev = timeline.adapt_cth_comments(comments, kind="issue", number=637)[0] + self.assertEqual(ev.event_type, timeline.UNKNOWN_HANDOFF_EVENT_TYPE) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(ev.to_dict())) + + def test_cth_unknown_and_whitespace_headings_normalized(self): + for heading in ("Totally Made Up", "Author Handoff", "author handoff"): + with self.subTest(heading=heading): + comments = [{ + "id": 13, + "body": _raw_cth_body(heading), + "created_at": "2026-07-23T05:00:00Z", + "user": {"login": "x"}, + }] + events = timeline.adapt_cth_comments(comments, kind="issue", number=637) + self.assertEqual(len(events), 1) + self.assertEqual(events[0].event_type, timeline.UNKNOWN_HANDOFF_EVENT_TYPE) + + def test_cth_declared_types_all_preserved(self): + """Point 5: no legitimate declared type is lost to the new check.""" + for cth_type in sorted(CTH_TYPES): + with self.subTest(cth_type=cth_type): + comments = [{ + "id": 14, + "body": _raw_cth_body(cth_type), + "created_at": "2026-07-23T05:00:00Z", + "user": {"login": "x"}, + }] + ev = timeline.adapt_cth_comments(comments, kind="issue", number=637)[0] + self.assertEqual(ev.event_type, f"handoff:{cth_type}") + self.assertFalse(ev.sensitive) + + def test_cth_surrounding_whitespace_still_matches_contract(self): + comments = [{ + "id": 15, + "body": _raw_cth_body(" Author Handoff "), + "created_at": "2026-07-23T05:00:00Z", + "user": {"login": "x"}, + }] + ev = timeline.adapt_cth_comments(comments, kind="issue", number=637)[0] + self.assertEqual(ev.event_type, "handoff:Author Handoff") + + # ---- complete serialized payload, both paths -------------------------- # + + def _payload_clean(self, snapshot): + payload = json.dumps(snapshot.to_dict()) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, payload) + from webui import console_redaction + + self.assertEqual(console_redaction.scan_for_secrets(snapshot.to_dict()), []) + return snapshot.to_dict() + + def test_canary_absent_from_full_payload_via_cp_event_type(self): + import tempfile + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + _seed_event( + db, + event_type=SYNTHETIC_SECRET_40_HEX, + message="routine deploy", + created_at="2026-07-23T06:00:00Z", + ) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, db_path=db, comment_source=lambda k, n: [], + ) + d = self._payload_clean(snap) + types = [e["event_type"] for e in d["events"]] + self.assertIn(timeline.UNSAFE_EVENT_TYPE, types) + # The legitimate seeded types are still present and unchanged. + self.assertIn("allocation", types) + + def test_canary_absent_from_full_payload_via_cth_heading(self): + import tempfile + + comments = [{ + "id": 21, + "body": _raw_cth_body(SYNTHETIC_SECRET_40_HEX), + "created_at": "2026-07-23T09:00:00Z", + "user": {"login": "jcwalker3"}, + }] + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, db_path=db, comment_source=lambda k, n: comments, + ) + d = self._payload_clean(snap) + self.assertIn( + timeline.UNKNOWN_HANDOFF_EVENT_TYPE, + [e["event_type"] for e in d["events"]], + ) + + def test_canary_absent_when_seeded_through_both_paths_at_once(self): + import tempfile + + comments = [{ + "id": 22, + "body": _raw_cth_body(SYNTHETIC_SECRET_40_HEX), + "created_at": "2026-07-23T09:00:00Z", + "user": {"login": "jcwalker3"}, + }] + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + _seed_event( + db, + event_type=SYNTHETIC_SECRET_40_HEX, + message=f"deploy token={SYNTHETIC_SECRET_40_HEX}", + created_at="2026-07-23T06:00:00Z", + ) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, db_path=db, comment_source=lambda k, n: comments, + ) + self._payload_clean(snap) + + +class TestSerializedFieldAudit(unittest.TestCase): + """The remaining externally influenced strings that reach the payload.""" + + def test_event_key_ids_must_be_plain_identifiers(self): + rows = [ + {"event_id": SYNTHETIC_SECRET_40_HEX, "event_type": "allocation", + "message": "m", "created_at": "2026-07-23T01:00:00Z", + "kind": "issue", "number": 637}, + {"event_id": 8, "event_type": "allocation", "message": "m", + "created_at": "2026-07-23T01:00:00Z", "kind": "issue", "number": 637}, + ] + events = timeline.adapt_cp_events(rows) + self.assertEqual([e.event_key for e in events], ["cp:8"]) + + def test_comment_id_must_be_a_plain_identifier(self): + comments = [{ + "id": SYNTHETIC_SECRET_40_HEX, + "body": _raw_cth_body("Author Handoff"), + "created_at": "2026-07-23T05:00:00Z", + "user": {"login": "x"}, + }] + self.assertEqual(timeline.adapt_cth_comments(comments, kind="issue", number=637), []) + + def test_adapter_refuses_a_scope_it_cannot_express(self): + comments = [{ + "id": 1, + "body": _raw_cth_body("Author Handoff"), + "created_at": "2026-07-23T05:00:00Z", + "user": {"login": "x"}, + }] + self.assertEqual( + timeline.adapt_cth_comments(comments, kind="issue", number=SYNTHETIC_SECRET_40_HEX), + [], + ) + self.assertEqual( + timeline.adapt_cth_comments(comments, kind=SYNTHETIC_SECRET_40_HEX, number=1), + [], + ) + + def test_source_failure_reason_is_redacted(self): + def boom(kind, number): + raise RuntimeError(f"auth failed with token={SYNTHETIC_SECRET_40_HEX}") + + import tempfile + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + snap = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, db_path=db, comment_source=boom, + ) + d = snap.to_dict() + handoff = [s for s in d["sources"] if s["name"] == "gitea_handoff"][0] + self.assertFalse(handoff["ok"]) + self.assertIn("handoff source failed", handoff["reason"]) + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(d)) + + def test_echoed_scope_and_filters_are_guarded(self): + import tempfile + + with tempfile.TemporaryDirectory() as tmp: + db = os.path.join(tmp, "cp.sqlite3") + _seed_db(db) + d = timeline.load_timeline( + remote="prgs", org="Scaled-Tech-Consulting", repo="Gitea-Tools", + issue=637, session=SYNTHETIC_SECRET_40_HEX, db_path=db, + comment_source=lambda k, n: [], + ).to_dict() + self.assertNotIn(SYNTHETIC_SECRET_40_HEX, json.dumps(d)) + # Ordinary scope values are untouched, so the echo stays useful. + self.assertEqual( + d["scope"], + {"remote": "prgs", "org": "Scaled-Tech-Consulting", "repo": "Gitea-Tools"}, + ) + self.assertEqual(d["filters"]["issue"], 637) + + +class TestCthTypeContract(unittest.TestCase): + """The contract has one authority; the read path consults it.""" + + def test_declared_types_are_known(self): + for cth_type in CTH_TYPES: + self.assertTrue(is_known_cth_type(cth_type)) + + def test_undeclared_types_are_not_known(self): + for value in ("", None, "Made Up", SYNTHETIC_SECRET_40_HEX, "author handoff"): + self.assertFalse(is_known_cth_type(value)) + + def test_parse_reports_contract_membership(self): + known = parse_cth_comment(_raw_cth_body("Author Handoff")) + self.assertTrue(known["cth_type_known"]) + self.assertEqual(known["cth_type"], "Author Handoff") + unknown = parse_cth_comment(_raw_cth_body("Not A Real Type")) + # Parsing stays total: the type is still reported, just not endorsed. + self.assertEqual(unknown["cth_type"], "Not A Real Type") + self.assertFalse(unknown["cth_type_known"]) + + +if __name__ == "__main__": + unittest.main() diff --git a/tests/test_worktree_cleanup_audit.py b/tests/test_worktree_cleanup_audit.py index f55b652..f3fc510 100644 --- a/tests/test_worktree_cleanup_audit.py +++ b/tests/test_worktree_cleanup_audit.py @@ -134,13 +134,35 @@ class TestClassification(unittest.TestCase): self.assertEqual(cls, wca.CLASS_ACTIVE_OPEN_PR) self.assertFalse(wca.is_removable(cls)) - def test_stale_clean_issue_worktree_removable(self): - # Scenario 5: clean issue worktree, TTL expired, no lock -> removable. + def test_stale_clean_issue_worktree_needs_merged_pr_proof(self): + # Scenario 5 (#858): age is not proof that the branch landed, so a + # TTL-expired issue worktree stays active work. Only authoritative + # merged-PR evidence makes it removable, which is what keeps a + # worktree holding unmerged commits from being reclaimed by age. cls = wca.classify_worktree( workflow_type=wca.WORKFLOW_ISSUE_WORK, is_dirty=False, ttl_expired=True, ) + self.assertEqual(cls, wca.CLASS_ACTIVE_ISSUE_WORK) + self.assertFalse(wca.is_removable(cls)) + + cls = wca.classify_worktree( + workflow_type=wca.WORKFLOW_ISSUE_WORK, + is_dirty=False, + ttl_expired=True, + merged_pr_cleanup={"proven": True}, + ) + self.assertEqual(cls, wca.CLASS_CLEAN_STALE_REMOVABLE) + self.assertTrue(wca.is_removable(cls)) + + def test_stale_clean_conflict_fix_worktree_removable(self): + # conflict_fix keeps the original TTL rule; #858 changed issue work only. + cls = wca.classify_worktree( + workflow_type=wca.WORKFLOW_CONFLICT_FIX, + is_dirty=False, + ttl_expired=True, + ) self.assertEqual(cls, wca.CLASS_CLEAN_STALE_REMOVABLE) self.assertTrue(wca.is_removable(cls)) diff --git a/webui/app.py b/webui/app.py index 11b1a35..9ec5f17 100644 --- a/webui/app.py +++ b/webui/app.py @@ -12,6 +12,7 @@ from starlette.routing import Route from webui.deployment_boundary import deployment_snapshot from webui.layout import render_page +from webui.nav import NAV_GROUPS, STUB_PAGES from webui.project_registry import ( ProjectRegistry, RegistryError, @@ -47,12 +48,14 @@ from webui.runtime_health import load_runtime_snapshot, snapshot_to_dict as runt from webui.runtime_views import render_runtime_page from webui.policy_inventory import load_policy_inventory, snapshot_to_dict as policy_snapshot_to_dict from webui.policy_views import render_policy_page +from webui.timeline import load_timeline, snapshot_to_dict as timeline_snapshot_to_dict from webui.system_health import ( API_PATH as SYSTEM_HEALTH_API_PATH, load_system_health, process_uptime, snapshot_to_dict as system_health_to_dict, ) +from webui.system_health_views import render_system_health_page _READ_ONLY_METHODS = frozenset({"GET", "HEAD", "OPTIONS"}) _AUDIT_MUTATION_PATHS = frozenset({"/audit", "/api/audit"}) @@ -67,25 +70,62 @@ def _stub_page(title: str, description: str) -> HTMLResponse: return HTMLResponse(render_page(title=title, body_html=body)) +_LEGACY_PAGES = ( + ("/queue", "Queue", "live PR and issue dashboard (#429)"), + ("/projects", "Projects", "registry and onboarding (#427)"), + ("/prompts", "Prompts", "canonical workflow prompt library (#428)"), + ("/runtime", "Runtime", "MCP health and stale-runtime detection (#430)"), + ("/audit", "Audit", "final-report paste and validator preview (#431)"), + ("/worktrees", "Worktrees", "branch hygiene dashboard (#432)"), + ("/leases", "Leases", "collision and lease visibility (#433)"), + ("/actions", "Actions", "gated write-action framework (#434)"), +) + + +def _render_home_nav_groups() -> str: + groups = [] + for group in NAV_GROUPS: + items = "".join( + f'
  • {item.label}' + + ("" if item.status == "live" else " (stub)") + + "
    • " + for item in group.items + ) + groups.append(f"

    {group.label}

      {items}
    ") + return "".join(groups) + + async def home(_request: Request) -> HTMLResponse: + legacy = "".join( + f"
  • {label} — {desc} " + f'({href})
    • ' + for href, label, desc in _LEGACY_PAGES + ) body = ( "

    Operator console

    " - "

    Local entry point for MCP Control Plane operational views.

    " - "
      " - "
    • Queue — live PR and issue dashboard (#429)
      • " - "
    • Projects — registry and onboarding (#427)
      • " - "
    • Prompts — canonical workflow prompt library (#428)
      • " - "
    • Runtime — MCP health and stale-runtime detection (#430)
      • " - "
    • Policy — workflow guardrail configuration visibility (#646)
      • " - "
    • Audit — final-report paste and validator preview (#431)
      • " - "
    • Worktrees — branch hygiene dashboard (#432)
      • " - "
    • Leases — collision and lease visibility (#433)
      • " - "
    • Actions — gated write-action framework (#434)
      • " - "
    " + "

    Read-only home for the MCP Control Plane Phase 1 operator console. " + "Gitea, MCP capability gates, and canonical workflows remain the source " + "of truth; this console never mutates them.

    " + "

    Phase 1 surfaces

    " + + _render_home_nav_groups() + + "

    MVP legacy pages

    " + "
      " + legacy + "
    " ) return HTMLResponse(render_page(title="Home", body_html=body)) +async def phase_stub(request: Request) -> HTMLResponse: + """Graceful read-only placeholder for a not-yet-implemented Phase 1 surface.""" + title, description = STUB_PAGES[request.url.path] + body = ( + f"

    {title}

    " + f'

    {description}

    ' + "

    Phase 1 shell placeholder — no write actions. Tracked under " + "epic #631.

    " + ) + return HTMLResponse(render_page(title=title, body_html=body)) + + async def health(_request: Request) -> JSONResponse: """Liveness only — deliberately cheap, runs no dependency probe (#634). @@ -124,6 +164,24 @@ async def api_system_health(request: Request) -> JSONResponse: return JSONResponse(payload, status_code=200 if snapshot.ready else 503) +async def system_health(request: Request) -> HTMLResponse: + """Read-only system-health dashboard (#639). + + Shares the #634 snapshot loader with the JSON API so the page can never + disagree with it. `?deep=1` opts into the network probe exactly as the API + does; the default page load stays cheap. The response is always 200: this + is an operator view that must render the degraded state, not withhold it. + """ + deep = _truthy_flag(request.query_params.get("deep")) + snapshot = load_system_health(deep=deep) + return HTMLResponse( + render_page( + title="System health", + body_html=render_system_health_page(snapshot), + ) + ) + + async def queue(_request: Request) -> HTMLResponse: snapshot = load_queue_snapshot() return HTMLResponse(render_page(title="Queue", body_html=render_queue_page(snapshot))) @@ -424,6 +482,104 @@ async def api_console_security_model(_request: Request) -> JSONResponse: }) +def _query_int(request: Request, key: str) -> int | None: + """Parse an optional integer query parameter; None when absent/invalid.""" + raw = request.query_params.get(key) + if raw is None or not str(raw).strip(): + return None + try: + return int(str(raw).strip()) + except (TypeError, ValueError): + return None + + +def _derive_remote(host: str) -> str: + """Map a Gitea host to its known short remote name (control-plane scope key).""" + text = (host or "").lower() + if "prgs" in text: + return "prgs" + if "dadeschools" in text: + return "dadeschools" + return text.split(".")[0] if text else "" + + +def _timeline_comment_source(host: str, org: str, repo: str): + """Build a fail-soft CTH-comment fetcher for one repo, or None when offline. + + Returns a callable ``(kind, number) -> list[comment]``. Credentials or + network failures raise inside the callable so ``load_timeline`` degrades the + handoff source rather than the whole timeline. Offline test mode yields no + live source so the handoff section reports ``not run``. + """ + import os + + from gitea_auth import api_fetch_page, get_auth_header, repo_api_url + + offline = (os.environ.get("WEBUI_TEST_OFFLINE") or "").strip().lower() in {"1", "true", "yes"} + if offline: + return None + auth = get_auth_header(host) + if not auth: + return None + + def _fetch(kind: str, number: int) -> list: + segment = "pulls" if kind == "pr" else "issues" + url = f"{repo_api_url(host, org, repo)}/{segment}/{int(number)}/comments" + comments: list = [] + page = 1 + while page <= 20: + raw, meta = api_fetch_page(url, auth, page=page, limit=50) + comments.extend(raw) + if bool(meta["is_final_page"]): + break + page += 1 + return comments + + return _fetch + + +async def api_v1_timeline(request: Request) -> JSONResponse: + """Read-only workflow-event timeline (#637). Filter by issue/PR/session.""" + from webui.queue_loader import _host_from_url # host normalisation helper + + registry, error = _load_project_registry() + if error is not None: + return JSONResponse(error.to_dict(), status_code=500) + project = registry.projects[0] if registry.projects else None + + org = request.query_params.get("org") or (project.gitea_owner if project else "") + repo = request.query_params.get("repo") or (project.repo_name if project else "") + host = _host_from_url(project.remote_host) if project else "" + remote = request.query_params.get("remote") or _derive_remote(host) + + if not (remote and org and repo): + return JSONResponse( + { + "error": "timeline_scope_unresolved", + "detail": "no project in registry and no remote/org/repo query params provided", + }, + status_code=400, + ) + + comment_source = _timeline_comment_source(host, org, repo) if (host and org and repo) else None + + snapshot = load_timeline( + remote=remote, + org=org, + repo=repo, + issue=_query_int(request, "issue"), + pr=_query_int(request, "pr"), + session=(request.query_params.get("session") or None), + limit=_query_int(request, "limit"), + offset=_query_int(request, "offset"), + comment_source=comment_source, + ) + # A filter no surviving source can carry is refused, not answered empty: + # a 200 with zero events would tell the operator no such activity exists. + status_code = 200 if snapshot.ok else 422 + return JSONResponse(timeline_snapshot_to_dict(snapshot), status_code=status_code) + + async def method_not_allowed(request: Request, _exc: Exception) -> Response: path = request.url.path if path in _AUDIT_MUTATION_PATHS and request.method == "POST": @@ -447,6 +603,7 @@ def create_app(*, bind_host: str | None = None) -> Starlette: Route("/", home, methods=["GET"]), Route("/health", health, methods=["GET"]), Route(SYSTEM_HEALTH_API_PATH, api_system_health, methods=["GET"]), + Route("/system-health", system_health, methods=["GET"]), Route("/queue", queue, methods=["GET"]), Route("/api/queue", api_queue, methods=["GET"]), Route("/projects", projects, methods=["GET"]), @@ -465,6 +622,7 @@ def create_app(*, bind_host: str | None = None) -> Starlette: Route("/api/runtime", api_runtime, methods=["GET"]), Route("/policy", policy, methods=["GET"]), Route("/api/v1/policy", api_v1_policy, methods=["GET"]), + Route("/api/v1/timeline", api_v1_timeline, methods=["GET"]), Route("/audit", audit, methods=["GET", "POST"]), Route("/api/audit", api_audit, methods=["GET", "POST"]), Route("/worktrees", worktrees, methods=["GET"]), @@ -488,6 +646,10 @@ def create_app(*, bind_host: str | None = None) -> Starlette: api_console_security_model, methods=["GET"], ), + *[ + Route(path, phase_stub, methods=["GET"]) + for path in STUB_PAGES + ], ], exception_handlers={405: method_not_allowed}, ) diff --git a/webui/layout.py b/webui/layout.py index 3adf375..6d8226c 100644 --- a/webui/layout.py +++ b/webui/layout.py @@ -2,29 +2,66 @@ from __future__ import annotations -NAV_ITEMS = ( - ("/", "Home"), - ("/queue", "Queue"), - ("/projects", "Projects"), - ("/prompts", "Prompts"), - ("/runtime", "Runtime"), - ("/policy", "Policy"), - ("/audit", "Audit"), - ("/worktrees", "Worktrees"), - ("/leases", "Leases"), - ("/actions", "Actions"), -) +import os + +from webui.nav import NAV_GROUPS MVP_NOTICE = ( "Read-only MVP — Gitea, MCP tools, and canonical workflows remain the " "source of truth. No mutation endpoints." ) +# Canonical docs entry point surfaced from the shell header (#638). +DOCS_URL = ( + "https://gitea.prgs.cc/Scaled-Tech-Consulting/Gitea-Tools/src/branch/" + "master/docs/webui-local-dev.md" +) + +_LOCAL_HOSTS = frozenset({"", "127.0.0.1", "localhost", "::1"}) + + +def environment_label() -> str: + """Classify the serving environment as ``local`` or ``remote`` (#638). + + Derived from the same ``WEBUI_HOST`` default the app binds to; loopback + hosts are ``local``, anything else is ``remote``. Read-only signal only. + """ + host = (os.environ.get("WEBUI_HOST", "127.0.0.1") or "").strip().lower() + return "local" if host in _LOCAL_HOSTS else "remote" + + +def _render_nav() -> str: + groups_html = [] + for group in NAV_GROUPS: + links = "".join( + f'{item.label}' + for item in group.items + ) + groups_html.append( + '
    ' + f'{group.label}' + f'{links}' + "
    " + ) + return "".join(groups_html) + + +def _render_badges() -> str: + env = environment_label() + return ( + '
    ' + f'env: {env}' + 'mode: read-only' + f'Docs' + "
    " + ) + def render_page(*, title: str, body_html: str, extra_head: str = "") -> str: - nav_links = "".join( - f'{label}' for href, label in NAV_ITEMS - ) + nav_links = _render_nav() + header_badges = _render_badges() return f""" @@ -54,21 +91,58 @@ def render_page(*, title: str, body_html: str, extra_head: str = "") -> str: padding: 0.75rem 1.25rem; }} header h1 {{ - margin: 0 0 0.5rem; + margin: 0; font-size: 1.1rem; font-weight: 600; }} + .header-top {{ + display: flex; + flex-wrap: wrap; + align-items: center; + justify-content: space-between; + gap: 0.5rem 1rem; + margin-bottom: 0.6rem; + }} + .header-badges {{ display: inline-flex; flex-wrap: wrap; gap: 0.4rem; }} + .env-badge.env-local {{ color: #8fd19e; border-color: #3d6b4a; }} + .env-badge.env-remote {{ color: #e0c27a; border-color: #6b5730; }} + .mode-badge {{ color: #9ec8f0; border-color: #3d5f7a; }} + a.docs-link {{ + color: var(--accent); + border-color: var(--accent); + text-decoration: none; + text-transform: none; + }} + a.docs-link:hover {{ filter: brightness(1.12); }} nav {{ display: flex; flex-wrap: wrap; - gap: 0.75rem 1rem; + gap: 0.5rem 1.25rem; }} + .nav-group {{ + display: flex; + flex-direction: column; + gap: 0.15rem; + }} + .nav-group-label {{ + font-size: 0.68rem; + text-transform: uppercase; + letter-spacing: 0.04em; + color: var(--muted); + }} + .nav-group-links {{ display: inline-flex; flex-wrap: wrap; gap: 0.6rem; }} nav a {{ color: var(--accent); text-decoration: none; font-size: 0.9rem; }} nav a:hover {{ text-decoration: underline; }} + nav a.nav-stub {{ color: var(--muted); }} + nav a.nav-stub::after {{ + content: " ·stub"; + font-size: 0.7rem; + color: var(--muted); + }} main {{ max-width: 52rem; margin: 0 auto; @@ -162,12 +236,34 @@ def render_page(*, title: str, body_html: str, extra_head: str = "") -> str: .badge-in-review {{ color: #9ec8f0; border-color: #3d5f7a; }} .badge-duplicate {{ color: #e0c27a; border-color: #6b5730; }} .badge-stale {{ color: #c9b8e8; border-color: #5a4a78; }} + .badge-health-ok {{ color: #8fd19e; border-color: #3d6b4a; }} + .badge-health-degraded {{ color: #e0c27a; border-color: #6b5730; }} + .badge-health-down {{ color: #f0a8a8; border-color: #7a3b3b; }} + .badge-health-skipped {{ color: var(--muted); }} + .badge-health-unproven {{ color: #c9b8e8; border-color: #5a4a78; }} + .health-card {{ + margin: 1.25rem 0; + padding: 0.85rem 1rem 1rem; + border: 1px solid var(--border); + border-radius: 8px; + background: var(--surface); + }} + .health-card h3 {{ margin: 0 0 0.5rem; font-size: 1.05rem; }} + .health-card h4 {{ margin: 1rem 0 0.35rem; font-size: 0.92rem; color: var(--muted); }} + .health-headline {{ color: var(--text); font-size: 1rem; margin: 0 0 0.5rem; }} + .health-degraded {{ border-left-color: #e0c27a; }} + .health-stale {{ border-left-color: #f0a8a8; }} + ul.reasons {{ margin: 0.35rem 0; padding-left: 1.15rem; color: var(--muted); font-size: 0.9rem; }} + ul.reasons li {{ margin-bottom: 0.3rem; }} {extra_head}
    -

    MCP Control Plane

    +
    +

    MCP Control Plane

    + {header_badges} +
    diff --git a/webui/nav.py b/webui/nav.py new file mode 100644 index 0000000..c24643d --- /dev/null +++ b/webui/nav.py @@ -0,0 +1,112 @@ +"""Navigation IA for the Phase 1 operator console shell (#638). + +Single source of truth for the console navigation so ``webui/layout.py`` and +the ``webui/app.py`` route table stay aligned with epic #631. Read-only: every +destination is a GET view or a Phase 1 placeholder. No mutation links. + +Nav groups follow the #631 Phase 1 information architecture: Health, Traffic, +Runtime/Sessions, Projects, Inventory, Timeline, Policy (placeholder), and +Insights (placeholder). Later-phase surfaces are declared as ``stub`` items and +backed by ``STUB_PAGES`` so their nav links resolve to a graceful placeholder +instead of a 404. +""" + +from __future__ import annotations + +from dataclasses import dataclass + + +@dataclass(frozen=True) +class NavItem: + """A single navigation destination. + + ``status`` is ``"live"`` for implemented views and ``"stub"`` for Phase 1 + placeholders whose backing view lands in a later child issue. + """ + + href: str + label: str + status: str = "live" + + +@dataclass(frozen=True) +class NavGroup: + label: str + items: tuple[NavItem, ...] + + +NAV_GROUPS: tuple[NavGroup, ...] = ( + NavGroup("Health", ( + NavItem("/health", "Liveness"), + NavItem("/system-health", "System health"), + )), + NavGroup("Traffic", ( + NavItem("/queue", "Queue"), + NavItem("/leases", "Leases"), + NavItem("/actions", "Actions"), + )), + NavGroup("Runtime/Sessions", ( + NavItem("/runtime", "Runtime health"), + NavItem("/sessions", "Sessions", "stub"), + )), + NavGroup("Projects", ( + NavItem("/projects", "Projects"), + )), + NavGroup("Inventory", ( + NavItem("/inventory", "Inventory", "stub"), + NavItem("/worktrees", "Worktrees"), + )), + NavGroup("Timeline", ( + NavItem("/timeline", "Timeline", "stub"), + )), + NavGroup("Policy", ( + NavItem("/policy", "Policy", "stub"), + NavItem("/prompts", "Prompts"), + )), + NavGroup("Insights", ( + NavItem("/insights", "Insights", "stub"), + NavItem("/audit", "Audit"), + )), +) + + +# Phase 1 placeholder destinations whose backing views land in later child +# issues of epic #631. Each maps a path to (title, description). Routes are +# registered so nav links resolve to a graceful, read-only stub page. +STUB_PAGES: dict[str, tuple[str, str]] = { + "/sessions": ( + "Sessions", + "Active session, capability, and role inventory. Backed by the unified " + "inventory API (#636) once it lands.", + ), + "/inventory": ( + "Inventory", + "Unified sessions, leases, locks, namespaces, and worktree inventory. " + "Backed by the Phase 1 inventory API (#636).", + ), + "/timeline": ( + "Timeline", + "Workflow event timeline across issues and PRs. A later Phase 1 surface.", + ), + "/policy": ( + "Policy", + "Capability and role policy surface. Placeholder until a later phase.", + ), + "/insights": ( + "Insights", + "Aggregate operational insights and trends. Placeholder until a later " + "phase.", + ), +} + + +def iter_nav_items(): + """Yield every ``NavItem`` across all groups in declared order.""" + for group in NAV_GROUPS: + for item in group.items: + yield item + + +def nav_hrefs() -> tuple[str, ...]: + """Return every navigation href in declared order.""" + return tuple(item.href for item in iter_nav_items()) diff --git a/webui/system_health_views.py b/webui/system_health_views.py new file mode 100644 index 0000000..e7ba492 --- /dev/null +++ b/webui/system_health_views.py @@ -0,0 +1,307 @@ +"""HTML views for the system-health dashboard (#639). + +Renders the read-only :class:`~webui.system_health.SystemHealthSnapshot` +produced by the Phase 1 system-health API (#634). The page offers no restart, +reload, or process-kill control: those are Phase 2 work, and manual process +kills are the contamination path #630 exists to prevent. + +Every free-text field passes through :func:`webui.system_health.redact` before +it reaches HTML, so a probe detail that captured a token or a credentialed URL +cannot leak through the dashboard even though the API redacts it already. +""" + +from __future__ import annotations + +import html + +from webui.system_health import ( + STATUS_DEGRADED, + STATUS_DOWN, + STATUS_OK, + STATUS_SKIPPED, + STATUS_UNPROVEN, + DependencyProbe, + SystemHealthSnapshot, + redact, +) + +_STATUS_BADGE_CLASS = { + STATUS_OK: "badge-health-ok", + STATUS_DEGRADED: "badge-health-degraded", + STATUS_DOWN: "badge-health-down", + STATUS_SKIPPED: "badge-health-skipped", + STATUS_UNPROVEN: "badge-health-unproven", +} + + +def _safe(value: object) -> str: + """Escape free text for HTML after redacting anything secret-shaped. + + Use this for every value that can carry arbitrary text — probe details, + reasons, probe errors — because those are where a credential could ride + along. + """ + return html.escape(redact(str(value))) + + +def _esc(value: object) -> str: + """Escape a structured field for HTML without redacting it. + + Commit SHAs, probe names, statuses, and timestamps are enumerated or + machine-generated, never credential-bearing. They must not go through + :func:`redact`: its opaque-token rule matches any 32-plus-character run, + so a 40-character git SHA would render as ``[redacted]`` and the parity + view — the one thing an operator reads this page for — would be blank. + """ + return html.escape(str(value)) + + +def _status_badge(status: str) -> str: + css = _STATUS_BADGE_CLASS.get(status, "badge-health-unproven") + return f'{_esc(status)}' + + +def _reason_list(reasons: tuple[str, ...], *, empty: str) -> str: + if not reasons: + return f"

    {html.escape(empty)}

    " + items = "".join(f"
  • {_safe(reason)}
    • " for reason in reasons) + return f"
      {items}
    " + + +def _readiness_card(snapshot: SystemHealthSnapshot) -> str: + """Overall readiness. + + ``ready`` and ``readiness_complete`` are shown separately on purpose: a + snapshot whose required probes never ran is not the same as one that ran + them and passed, and collapsing the two would render an unproven green. + """ + if snapshot.ready and snapshot.readiness_complete: + headline = "Ready" + elif snapshot.ready: + headline = "Ready (incomplete evidence)" + else: + headline = "Not ready" + + return ( + "
    " + f"

    Readiness {_status_badge(snapshot.status)}

    " + f"

    {html.escape(headline)}

    " + "" + f"" + f"" + f"" + "" + f"" + "" + f"" + f"" + "
    Service{_esc(snapshot.service)}
    Mode{_esc(snapshot.mode)}
    Ready{_esc(snapshot.ready)}
    Readiness evidence complete{_esc(snapshot.readiness_complete)}
    Deep probes requested{_esc(snapshot.deep_probes_requested)}
    Observed at{_esc(snapshot.timestamp)}
    " + "

    Readiness reasons

    " + f"{_reason_list(snapshot.readiness_reasons, empty='No readiness objections recorded.')}" + "
    " + ) + + +def _version_card(snapshot: SystemHealthSnapshot) -> str: + version = snapshot.version + uptime_hours = snapshot.uptime_seconds / 3600.0 + known = ( + "resolved" + if version.known + else "unresolved — version fields could not be read from the checkout" + ) + schema = version.control_plane_schema_version + return ( + "
    " + "

    Version and uptime

    " + "" + f"" + "" + f"" + "" + f"" + f"" + f"" + f"" + "" + f"" + "
    Git SHA{_esc(version.git_sha or 'unknown')}
    Git describe{_esc(version.git_describe or 'unknown')}
    Control-plane schema{_esc(schema if schema is not None else 'unknown')}
    Python{_esc(version.python_version)}
    Version status{html.escape(known)}
    Started at{_esc(snapshot.started_at)}
    Uptime{snapshot.uptime_seconds:.3f}s ({uptime_hours:.2f}h)
    " + "
    " + ) + + +def _dependency_rows(probes: tuple[DependencyProbe, ...]) -> str: + if not probes: + return "

    No dependency probes were reported.

    " + rows = [] + for probe in probes: + latency = ( + f"{probe.latency_ms:.1f} ms" if probe.latency_ms is not None else "n/a" + ) + rows.append( + "" + f"{_esc(probe.name)}" + f"{_esc(probe.kind)}" + f"{_status_badge(probe.status)}" + f"{_esc('required' if probe.required else 'optional')}" + f"{html.escape(latency)}" + f"{_safe(probe.detail)}" + "" + ) + return ( + "" + "" + "" + "" + f"{''.join(rows)}
    DependencyKindStatusRequirementLatencyDetail
    " + ) + + +def _dependency_card(snapshot: SystemHealthSnapshot) -> str: + degraded = [probe for probe in snapshot.dependencies if probe.ran and not probe.healthy] + not_run = [probe for probe in snapshot.dependencies if not probe.ran] + + banner = "" + if degraded: + names = ", ".join(sorted(probe.name for probe in degraded)) + banner += ( + "

    Degraded dependencies: " + f"{_esc(names)}

    " + ) + if not_run: + names = ", ".join(sorted(probe.name for probe in not_run)) + banner += ( + "

    Not probed: " + f"{_esc(names)} — these contribute no evidence and are not counted " + "as healthy.

    " + ) + + return ( + "
    " + "

    Dependencies

    " + f"{banner}" + f"{_dependency_rows(snapshot.dependencies)}" + "

    Details are redacted at the API boundary and again " + "before rendering; credentials are never displayed.

    " + "
    " + ) + + +def _namespace_card(snapshot: SystemHealthSnapshot) -> str: + if not snapshot.mcp_namespaces: + body = "

    No MCP namespaces are declared.

    " + else: + rows = [] + for entry in snapshot.mcp_namespaces: + rows.append( + "" + f"{_esc(entry.get('namespace'))}" + f"{_esc(entry.get('required_tool'))}" + f"{_status_badge(str(entry.get('status') or STATUS_UNPROVEN))}" + f"{_esc(entry.get('ide_namespace_proven'))}" + f"{_safe(entry.get('reason'))}" + "" + ) + body = ( + "" + "" + "" + "" + f"{''.join(rows)}
    NamespaceRequired toolStatusIDE-provenReason
    " + ) + return ( + "
    " + "

    MCP namespaces

    " + f"{body}" + "

    The web process runs outside the IDE-managed MCP " + "client, so namespace health is reported as unproven rather than " + "guessed (#543).

    " + "
    " + ) + + +def _stale_runtime_card(snapshot: SystemHealthSnapshot) -> str: + stale = snapshot.stale_runtime + if stale.stale: + warning = ( + "

    Stale runtime: " + "the running code, the checkout, and the remote-tracking commit " + "disagree. Capability gates may be evaluating obsolete code — " + "do not treat this runtime as mutation-safe.

    " + ) + elif not stale.determinable: + warning = ( + "

    Staleness " + "indeterminate: parity could not be proven, so this " + "runtime is not reported as mutation-safe.

    " + ) + else: + warning = "" + + return ( + "
    " + "

    Stale-runtime parity

    " + f"{warning}" + "" + "" + f"" + "" + f"" + "" + f"" + f"" + f"" + f"" + "
    Daemon head{_esc(stale.daemon_head or 'unknown')}
    Checkout head{_esc(stale.checkout_head or 'unknown')}
    Remote head{_esc(stale.remote_head or 'unknown')}
    Stale{_esc(stale.stale)}
    Determinable{_esc(stale.determinable)}
    Mutation safe{_esc(stale.mutation_safe)}
    " + f"{_reason_list(stale.reasons, empty='Runtime, checkout, and remote agree.')}" + "
    " + ) + + +def _probe_error_card(snapshot: SystemHealthSnapshot) -> str: + if not snapshot.probe_errors: + return "" + return ( + "
    " + "

    Probe errors

    " + f"{_reason_list(snapshot.probe_errors, empty='')}" + "
    " + ) + + +def _recovery_card() -> str: + """Sanctioned recovery pointers only — never a manual process kill (#630).""" + return ( + "
    " + "

    Recovery

    " + "

    This dashboard is read-only. Restart and reload " + "controls arrive in Phase 2 (#642); until then recovery runs through " + "the sanctioned client reconnect / operator restart path.

    " + "
      " + "
    • Runtime and session view — active profile, " + "workflow hashes, and shell health.
      • " + "
    • Reconnect the MCP client from the IDE, then re-run the blocked " + "cycle. Never kill the daemon process manually: unmanaged kills are " + "recorded as runtime contamination (#630).
      • " + "
    • See docs/webui-local-dev.md for the documented " + "recovery sequence.
      • " + "
    " + "
    " + ) + + +def render_system_health_page(snapshot: SystemHealthSnapshot) -> str: + """Render the full system-health dashboard body.""" + return ( + "

    System health

    " + "

    Read-only view of the Phase 1 system-health API " + "(/api/v1/system/health). Reload this page to refresh; " + "nothing here polls or mutates on your behalf.

    " + f"{_readiness_card(snapshot)}" + f"{_stale_runtime_card(snapshot)}" + f"{_version_card(snapshot)}" + f"{_dependency_card(snapshot)}" + f"{_namespace_card(snapshot)}" + f"{_probe_error_card(snapshot)}" + f"{_recovery_card()}" + ) diff --git a/webui/timeline.py b/webui/timeline.py new file mode 100644 index 0000000..1ec8284 --- /dev/null +++ b/webui/timeline.py @@ -0,0 +1,906 @@ +"""Workflow-event and conversation timeline model (#637, Phase 1). + +Operators cannot browse a unified timeline of workflow events, decisions, +tool calls, and handoffs: the evidence is scattered across control-plane +events, Gitea canonical handoff comments, and local logs. This module defines +one durable, versioned event schema and per-source adapters that normalise +those scattered records into a single ``WorkflowEvent`` stream, plus a +read-only query layer (filter by issue / PR / session, stable ordering, +pagination) that the ``/api/v1/timeline`` route serves. + +Design rules honoured here: + +- **Read-only.** Sources are read; nothing is mutated. The control-plane + database is opened through a ``mode=ro`` URI so a missing or unwritable DB + degrades to a reason instead of creating directories or running migrations. +- **Fail-soft per source.** An unavailable source degrades to a status with a + reason rather than raising, and a source that could not run is never + rendered as an empty-and-healthy timeline. +- **Answerable filters only.** Each source declares which filter dimensions it + can actually answer. A filter dimension no source that ran can carry is + refused with an explicit reason rather than silently matching nothing: an + empty page from an unanswerable filter reads to an operator as "no such + activity", which is a different — and false — statement. +- **Redaction at the boundary, fail closed.** Every free-text field (event + messages, redacted tool arguments, decision/proof text) is run through the + console redaction policy before it leaves this module, and *before* any + structured value is derived from it — evidence references are extracted from + redacted text, then independently revalidated before serialization. An + unredactable value becomes the placeholder, and a value that cannot be proven + safe is dropped — an unredacted payload is never emitted, and a generation + error never drops raw data to a caller or a log. +- **Stable ordering.** Events sort by ``(timestamp, source_rank, event_key)`` + with a deterministic tiebreak, so pagination is stable across calls and + events with equal or missing timestamps keep a fixed order. + +Non-goals (from the issue): no full chat replay, no mutation of historical +events, no unredacted tool-argument storage. +""" + +from __future__ import annotations + +import re +import sqlite3 +from dataclasses import dataclass, replace +from datetime import datetime, timezone +from typing import Any, Callable, Iterable + +import control_plane_db +from webui import console_redaction + +# The schema is versioned so consumers can branch on shape. Bump on any +# breaking change to WorkflowEvent's serialized form. +TIMELINE_SCHEMA_VERSION = 1 + +# Known event sources and their deterministic ordering rank. When two events +# carry the same timestamp, the source rank breaks the tie before the +# per-source event key, so a control-plane event and a handoff comment minted +# in the same second always sort in a fixed order. +SOURCE_CONTROL_PLANE = "control_plane" +SOURCE_GITEA_HANDOFF = "gitea_handoff" +_SOURCE_RANK = { + SOURCE_CONTROL_PLANE: 0, + SOURCE_GITEA_HANDOFF: 1, +} + +# The filter dimensions the query layer accepts. +FILTER_ISSUE = "issue" +FILTER_PR = "pr" +FILTER_SESSION = "session" + +# Which dimensions each source can actually answer. This is a property of the +# underlying records, not of the query code: the control-plane ``events`` table +# is (event_id, work_item_id, event_type, message, created_at) and carries no +# session identity at all, so no control-plane event can ever match a session +# filter. A CTH handoff comment can declare its session as a field, so the +# handoff source answers all three. Filtering on a dimension the surviving +# sources cannot carry is refused in ``load_timeline`` rather than answered +# with an empty page. +_SOURCE_FILTER_SUPPORT: dict[str, tuple[str, ...]] = { + SOURCE_CONTROL_PLANE: (FILTER_ISSUE, FILTER_PR), + SOURCE_GITEA_HANDOFF: (FILTER_ISSUE, FILTER_PR, FILTER_SESSION), +} + +# Why a source cannot answer a dimension, for the refusal reason an operator reads. +_SOURCE_FILTER_LIMITS: dict[tuple[str, str], str] = { + (SOURCE_CONTROL_PLANE, FILTER_SESSION): ( + "control-plane events carry no session identity " + "(the events table has no session column)" + ), +} + +# A timestamp far in the future so events with no parseable timestamp sort +# last (after everything real) instead of first, without raising. +_MISSING_TS_SORT = "9999-12-31T23:59:59Z" + + +def _parse_ts(value: str | None) -> str | None: + """Normalise a timestamp to ``...Z`` UTC, or None when unparseable.""" + if not value: + return None + text = str(value).strip() + if not text: + return None + candidate = text[:-1] + "+00:00" if text.endswith("Z") else text + try: + parsed = datetime.fromisoformat(candidate) + except ValueError: + return None + if parsed.tzinfo is None: + parsed = parsed.replace(tzinfo=timezone.utc) + return parsed.astimezone(timezone.utc).replace(microsecond=0).isoformat().replace("+00:00", "Z") + + +def _redact(value: Any) -> Any: + """Redact a single free-text field, failing closed to the placeholder.""" + if value is None: + return None + return console_redaction.redact_text(str(value)) + + +@dataclass(frozen=True) +class WorkflowEvent: + """One normalised timeline event. + + Every field is optional except ``source``/``event_type``/``event_key`` + because sources carry different subsets. The class is frozen so an adapted + event is an immutable record; a consumer that needs a variant builds a new + one rather than mutating history. + """ + + source: str + event_type: str + event_key: str + timestamp: str | None = None + actor: str | None = None + role: str | None = None + issue_number: int | None = None + pr_number: int | None = None + session_id: str | None = None + tool_name: str | None = None + decision: str | None = None + message: str | None = None + correlation_id: str | None = None + evidence_refs: tuple[str, ...] = () + sensitive: bool = False + + def sort_key(self) -> tuple[str, int, str]: + return ( + self.timestamp or _MISSING_TS_SORT, + _SOURCE_RANK.get(self.source, 99), + self.event_key, + ) + + def to_dict(self) -> dict[str, Any]: + return { + "source": self.source, + "event_type": self.event_type, + "event_key": self.event_key, + "timestamp": self.timestamp, + "actor": self.actor, + "role": self.role, + "issue_number": self.issue_number, + "pr_number": self.pr_number, + "session_id": self.session_id, + "tool_name": self.tool_name, + "decision": self.decision, + "message": self.message, + "correlation_id": self.correlation_id, + "evidence_refs": list(self.evidence_refs), + "sensitive": self.sensitive, + } + + +# --------------------------------------------------------------------------- # +# Adapters — pure functions from a source's raw records to WorkflowEvents. # +# Each is total: a malformed record is skipped, never raised on. # +# --------------------------------------------------------------------------- # + +# Event types whose payload is treated as sensitive and always redaction-hard +# (they can carry lease/session provenance or tool arguments). +_SENSITIVE_EVENT_HINTS = ("lease", "capability", "token", "auth", "secret") + +# Reference tokens (issue/PR/comment ids) and SHAs parsed out of proof text. +_EVIDENCE_REF_RE = re.compile(r"(?:#|PR\s*#?|issue\s*#?|comment\s*#?)(\d+)", re.IGNORECASE) + +# A commit reference is only recognised when the text *declares* it as one. +# A bare lowercase hex run is not evidence of anything: at 40 characters it is +# exactly the shape of a Gitea personal access token, and at 7 it also matches +# ordinary words such as "defaced". Requiring an anchoring keyword keeps real +# references ("commit abc1234", "at head a209756...", "base caaae9b6") usable +# while refusing to lift an undeclared secret-shaped run out of free text. +_SHA_RE = re.compile( + r"(?i:\b(?:commit|sha|head|base|parent|revision|rev|merge[- ]base)\b[\s:=@#]*)" + r"([0-9a-f]{7,40})\b" +) + +# Shapes a serialized evidence reference is allowed to take. Anything else is +# dropped rather than emitted. +_REF_ISSUE_SHAPE = re.compile(r"^#[0-9]{1,9}$") +_REF_SHA_SHAPE = re.compile(r"^[0-9a-f]{7,40}$") + +# A long undelimited hex run with no declaring context is treated as credential +# material wherever it appears, never as an identifier. +_BARE_SECRET_SHAPE = re.compile(r"^[0-9a-f]{32,}$") + +# An event type reads like an identifier, but a stored one is externally +# influenced: any producer that writes the control-plane ``events`` table +# chooses the string. It reaches ``to_dict`` verbatim, so it is validated here +# rather than trusted because of where it came from. +_CP_EVENT_TYPE_SHAPE = re.compile(r"^[A-Za-z][A-Za-z0-9._:+-]{0,63}$") + +# Emitted in place of a value that cannot be proven safe. Deliberately not a +# plausible workflow type: an unsafe value is refused, never quietly rewritten +# into a different valid-looking one that would misdescribe the record. +UNSAFE_EVENT_TYPE = "unsafe:redacted" + +# Emitted for a CTH heading that is not a declared member of ``CTH_TYPES``. The +# contract is enforced on write (``format_cth_body``) and on assess; the read +# path the timeline uses enforces it too rather than assuming it was. +UNKNOWN_HANDOFF_EVENT_TYPE = "handoff:unrecognized" + +# A source record id is a plain integer in both sources it comes from: the +# control-plane ``events`` primary key and a Gitea comment id. ``event_key`` is +# serialized verbatim and is the pagination tiebreak, so anything else is +# refused rather than interpolated into it. +_RECORD_ID_SHAPE = re.compile(r"^[0-9]{1,19}$") + + +def _kind_to_numbers(kind: str | None, number: int | None) -> tuple[int | None, int | None]: + """Map a control-plane work-item (kind, number) to (issue_no, pr_no).""" + if number is None: + return (None, None) + if kind == "pr": + return (None, int(number)) + if kind == "issue": + return (int(number), None) + return (None, None) + + +def _correlation_for(kind: str | None, number: int | None) -> str | None: + if number is None or kind not in ("issue", "pr"): + return None + return f"{kind}#{number}" + + +def _extract_evidence_refs(*texts: str | None) -> tuple[str, ...]: + """Extract issue/PR and declared-commit references from **redacted** text. + + Callers must pass text that has already been through :func:`_redact`; this + function derives a structured field from its input, so extracting ahead of + redaction would republish whatever redaction was about to remove. Every + reference is revalidated by :func:`_validated_evidence_refs` before it is + serialized. + """ + refs: list[str] = [] + for text in texts: + if not text: + continue + for match in _EVIDENCE_REF_RE.finditer(text): + token = f"#{match.group(1)}" + if token not in refs: + refs.append(token) + for match in _SHA_RE.finditer(text): + token = match.group(1) + if token not in refs: + refs.append(token) + return tuple(refs) + + +def _validated_evidence_refs(refs: Iterable[str]) -> tuple[tuple[str, ...], bool]: + """Independently revalidate references immediately before serialization. + + Extraction is not trusted on its own. A reference survives only when it has + a known reference shape and is unchanged by a second redaction pass — a + value the redaction policy would alter is credential material that must not + be emitted as a structured field. A full 40-character SHA stays usable + because extraction only accepts a hex run the source text explicitly + declared as a commit. Returns ``(safe_refs, dropped_any)``; ``dropped_any`` + marks the event sensitive so the drop is visible rather than silent. + """ + safe: list[str] = [] + dropped = False + for ref in refs or (): + try: + token = str(ref).strip() + if not token: + continue + recognised = bool(_REF_ISSUE_SHAPE.match(token) or _REF_SHA_SHAPE.match(token)) + if not recognised: + dropped = True + continue + if _redact(token) != token: + dropped = True + continue + if token not in safe: + safe.append(token) + except Exception: + # Fail closed: a reference that cannot be proven safe is dropped. + dropped = True + continue + return (tuple(safe), dropped) + + +def _safe_session_id(value: Any) -> str | None: + """Return a session identifier only when it is safe to emit. + + The value is authoritative source data — a session the record names for + itself — but it is still free text. It is dropped when redaction alters it + or when it is a bare secret-shaped hex run, so a credential parked in a + session field can never reach the payload or be echoed back by a filter. + """ + if value is None: + return None + text = str(value).strip() + if not text: + return None + if _BARE_SECRET_SHAPE.match(text): + return None + return text if _redact(text) == text else None + + +def _safe_record_id(value: Any) -> str | None: + """Return a source record id only when it is a plain numeric identifier. + + ``event_key`` is serialized verbatim and is the deterministic pagination + tiebreak, so an id is interpolated into it only when it has the shape both + real sources actually produce. A record whose identity cannot be trusted is + refused by the caller rather than keyed on. + """ + if value is None or isinstance(value, bool): + return None + if isinstance(value, int): + return str(value) + text = str(value).strip() + return text if _RECORD_ID_SHAPE.match(text) else None + + +def _safe_cp_event_type(value: Any) -> tuple[str, bool]: + """Validate a stored control-plane event type. Returns ``(type, unsafe)``. + + The stored value is externally influenced — whichever producer wrote the + ``events`` row chose the string — and ``to_dict`` serializes it verbatim, so + it passes a boundary of its own instead of relying on the one ``message`` + passes. A value survives only when it is an ordinary identifier, is not a + bare secret-shaped hex run, and is unchanged by a redaction pass. Anything + else fails closed to :data:`UNSAFE_EVENT_TYPE`: the record stays visible as + an audit entry, but the value itself is never republished — not verbatim, + not partially sanitized, and not rewritten into some other valid-looking + type that would misdescribe what happened. + """ + text = ("" if value is None else str(value)).strip() + if not text: + return ("", False) + if _BARE_SECRET_SHAPE.match(text): + return (UNSAFE_EVENT_TYPE, True) + if not _CP_EVENT_TYPE_SHAPE.match(text): + return (UNSAFE_EVENT_TYPE, True) + if _redact(text) != text: + return (UNSAFE_EVENT_TYPE, True) + return (text, False) + + +def _safe_echo(value: Any) -> Any: + """Guard a scalar that is echoed back rather than derived from a record. + + Query scope and filter values are caller-supplied and are reflected in the + response so an operator can see what was asked. Reflection is still + emission: a value redaction would alter, or a bare secret-shaped hex run, is + replaced by the placeholder instead of being echoed verbatim. Ordinary + scope and filter values pass through untouched. + """ + if value is None or isinstance(value, (int, bool)): + return value + text = str(value) + if _BARE_SECRET_SHAPE.match(text.strip()): + return console_redaction.REDACTED + return _redact(text) + + +def adapt_cp_events(rows: Iterable[dict[str, Any]]) -> list[WorkflowEvent]: + """Adapt control-plane ``events`` rows (joined to work_items) into events. + + Each row is expected to carry ``event_id``, ``event_type``, ``message``, + ``created_at`` and the joined work-item ``kind``/``number``. Rows missing + an id or type are skipped so a partially written table never raises. + """ + events: list[WorkflowEvent] = [] + for row in rows or []: + try: + event_id = _safe_record_id(row.get("event_id")) + raw_event_type = (row.get("event_type") or "").strip() + if event_id is None or not raw_event_type: + continue + # The stored type is source data, not a trusted constant: validate + # it before it is serialized, exactly as `message` below is redacted + # before it is serialized. + event_type, event_type_unsafe = _safe_cp_event_type(raw_event_type) + kind = row.get("kind") + number = row.get("number") + issue_no, pr_no = _kind_to_numbers(kind, number) + sensitive = event_type_unsafe or any( + hint in raw_event_type.lower() for hint in _SENSITIVE_EVENT_HINTS + ) + events.append( + WorkflowEvent( + source=SOURCE_CONTROL_PLANE, + event_type=event_type, + event_key=f"cp:{event_id}", + timestamp=_parse_ts(row.get("created_at")), + issue_number=issue_no, + pr_number=pr_no, + # No session_id: the control-plane events table is + # (event_id, work_item_id, event_type, message, created_at) + # and records no session. Inventing one from the work item + # or the message text would be a guess, so this source + # declares the session dimension unsupported instead + # (_SOURCE_FILTER_SUPPORT) and the query layer refuses a + # session filter it cannot honestly answer. + message=_redact(row.get("message")), + correlation_id=_correlation_for(kind, number), + sensitive=sensitive, + ) + ) + except Exception: + # A single malformed row must not sink the whole adaptation. + continue + return events + + +def adapt_cth_comments( + comments: Iterable[dict[str, Any]], + *, + kind: str, + number: int, +) -> list[WorkflowEvent]: + """Adapt Gitea Canonical Thread Handoff (CTH) comments into events. + + Only comments that parse as a CTH (``canonical_thread_handoff.parse_cth_comment``) + become events; ordinary comments are ignored. ``kind``/``number`` scope the + events to the issue or PR the comments belong to. + """ + # Imported lazily so this module has no import-time dependency on the + # handoff parser when only the control-plane adapter is used. + from canonical_thread_handoff import is_known_cth_type, parse_cth_comment + + # ``kind``/``number`` are interpolated into event_key and correlation_id, so + # they are normalised once here. A scope this adapter cannot express is + # refused outright rather than serialized into an identifier. + kind = (kind or "").strip().lower() + if kind not in ("issue", "pr"): + return [] + try: + number = int(number) + except (TypeError, ValueError): + return [] + + issue_no, pr_no = _kind_to_numbers(kind, number) + correlation = _correlation_for(kind, number) + events: list[WorkflowEvent] = [] + for comment in comments or []: + try: + body = comment.get("body") or "" + parsed = parse_cth_comment(body) + if not parsed: + continue + fields = parsed.get("fields") or {} + cth_type = parsed.get("cth_type") or "" + comment_id = _safe_record_id(comment.get("id")) + if comment_id is None: + continue + # The CTH heading is free text: the parser accepts whatever follows + # "## CTH:", and only the write and assess paths check it against + # the contract. Check it here too — an unrecognised heading is + # reported as such rather than serialized into event_type, so + # arbitrary, malformed, or secret-shaped heading content has no way + # through. Declared types are preserved exactly. + cth_type_known = is_known_cth_type(cth_type) + # Redaction runs first, and every derived value is taken from the + # redacted text — deriving evidence refs from the raw proof would + # re-emit exactly what redaction was about to remove. + decision = _redact(fields.get("decision")) + proof = _redact(fields.get("proof")) + next_action = _redact(fields.get("next action")) + refs, refs_dropped = _validated_evidence_refs( + _extract_evidence_refs(proof, decision) + ) + events.append( + WorkflowEvent( + source=SOURCE_GITEA_HANDOFF, + event_type=( + f"handoff:{cth_type.strip()}" + if cth_type_known + else UNKNOWN_HANDOFF_EVENT_TYPE + ), + event_key=f"cth:{kind}:{number}:{comment_id}", + timestamp=_parse_ts(comment.get("created_at")), + actor=_redact((comment.get("user") or {}).get("login")), + role=_redact(fields.get("next owner")), + issue_number=issue_no, + pr_number=pr_no, + # A CTH names its own session when the producer records one; + # it is read from that declared field, never inferred from + # unrelated text. + session_id=_safe_session_id(fields.get("session")), + decision=decision, + message=next_action or _redact(fields.get("status")), + correlation_id=correlation, + evidence_refs=refs, + sensitive=refs_dropped or not cth_type_known, + ) + ) + except Exception: + continue + return events + + +# --------------------------------------------------------------------------- # +# Read-only control-plane event source. # +# --------------------------------------------------------------------------- # + +_CP_EVENTS_QUERY = """ +SELECT e.event_id AS event_id, + e.event_type AS event_type, + e.message AS message, + e.created_at AS created_at, + w.kind AS kind, + w.number AS number +FROM events e +JOIN work_items w ON e.work_item_id = w.work_item_id +WHERE w.remote = ? AND w.org = ? AND w.repo = ? +""" + + +@dataclass(frozen=True) +class SourceStatus: + """Fail-soft status for one timeline source. + + ``supported_filters`` states which filter dimensions this source's records + can carry; ``unsupported_filters`` names the requested dimensions it cannot, + so an operator can see *why* a source contributed nothing rather than being + left to read an empty list as an absence of activity. + """ + + name: str + ok: bool + reason: str | None = None + count: int = 0 + supported_filters: tuple[str, ...] = () + unsupported_filters: tuple[str, ...] = () + + def to_dict(self) -> dict[str, Any]: + return { + "name": self.name, + "ok": self.ok, + "reason": self.reason, + "count": self.count, + "supported_filters": list(self.supported_filters), + "unsupported_filters": list(self.unsupported_filters), + } + + +def _cp_status(*, ok: bool, reason: str | None = None, count: int = 0) -> SourceStatus: + return SourceStatus( + SOURCE_CONTROL_PLANE, + ok=ok, + # A failure reason is serialized like any other field and is often an + # exception string carrying a path or a transport error, so it crosses + # the redaction boundary too. Static reasons pass through unchanged. + reason=_redact(reason), + count=count, + supported_filters=_SOURCE_FILTER_SUPPORT[SOURCE_CONTROL_PLANE], + ) + + +def _handoff_status(*, ok: bool, reason: str | None = None, count: int = 0) -> SourceStatus: + return SourceStatus( + SOURCE_GITEA_HANDOFF, + ok=ok, + # Same boundary as the control-plane status: this reason can quote an + # error raised by a live authenticated fetch. + reason=_redact(reason), + count=count, + supported_filters=_SOURCE_FILTER_SUPPORT[SOURCE_GITEA_HANDOFF], + ) + + +def read_cp_events( + *, + remote: str, + org: str, + repo: str, + db_path: str | None = None, +) -> tuple[list[WorkflowEvent], SourceStatus]: + """Read scoped control-plane events read-only. Never creates the DB. + + Opens the SQLite file through a ``mode=ro`` URI: a health/timeline read + must never create directories or run the schema migration that + ``ControlPlaneDB()`` performs on construction. A missing or unreadable DB + degrades to a status with a reason. + """ + path = (db_path or control_plane_db.default_db_path()).strip() + conn: sqlite3.Connection | None = None + try: + conn = sqlite3.connect(f"file:{path}?mode=ro", uri=True) + conn.row_factory = sqlite3.Row + cursor = conn.execute(_CP_EVENTS_QUERY, (remote, org, repo)) + rows = [dict(r) for r in cursor.fetchall()] + except sqlite3.OperationalError as exc: + return ([], _cp_status(ok=False, reason=f"control-plane DB unavailable: {exc}")) + except sqlite3.Error as exc: + return ([], _cp_status(ok=False, reason=f"control-plane read failed: {exc}")) + finally: + if conn is not None: + conn.close() + events = adapt_cp_events(rows) + return (events, _cp_status(ok=True, count=len(events))) + + +# --------------------------------------------------------------------------- # +# Filter, sort, paginate. # +# --------------------------------------------------------------------------- # + + +def filter_events( + events: Iterable[WorkflowEvent], + *, + issue: int | None = None, + pr: int | None = None, + session: str | None = None, +) -> list[WorkflowEvent]: + """Filter events by issue number, PR number, and/or session id. + + Filters are conjunctive. A filter that names a dimension an event does not + carry excludes that event (an issue filter excludes PR-only events). + """ + out: list[WorkflowEvent] = [] + for ev in events: + if issue is not None and ev.issue_number != issue: + continue + if pr is not None and ev.pr_number != pr: + continue + if session is not None and ev.session_id != session: + continue + out.append(ev) + return out + + +def sort_events(events: Iterable[WorkflowEvent]) -> list[WorkflowEvent]: + """Return events in stable timeline order (ascending).""" + return sorted(events, key=lambda ev: ev.sort_key()) + + +@dataclass(frozen=True) +class TimelinePage: + """One page of the sorted, filtered timeline.""" + + events: tuple[WorkflowEvent, ...] + total: int + limit: int + offset: int + + @property + def next_offset(self) -> int | None: + nxt = self.offset + len(self.events) + return nxt if nxt < self.total else None + + def to_dict(self) -> dict[str, Any]: + return { + "events": [ev.to_dict() for ev in self.events], + "pagination": { + "total": self.total, + "limit": self.limit, + "offset": self.offset, + "returned": len(self.events), + "next_offset": self.next_offset, + "has_more": self.next_offset is not None, + }, + } + + +_MAX_LIMIT = 500 +_DEFAULT_LIMIT = 50 + + +def _coerce_bounds(limit: int | None, offset: int | None) -> tuple[int, int]: + try: + lim = int(limit) if limit is not None else _DEFAULT_LIMIT + except (TypeError, ValueError): + lim = _DEFAULT_LIMIT + try: + off = int(offset) if offset is not None else 0 + except (TypeError, ValueError): + off = 0 + lim = max(1, min(lim, _MAX_LIMIT)) + off = max(0, off) + return (lim, off) + + +def paginate(events: list[WorkflowEvent], *, limit: int | None, offset: int | None) -> TimelinePage: + lim, off = _coerce_bounds(limit, offset) + window = events[off : off + lim] + return TimelinePage(events=tuple(window), total=len(events), limit=lim, offset=off) + + +# --------------------------------------------------------------------------- # +# Composition — load_timeline aggregates all sources, fail-soft. # +# --------------------------------------------------------------------------- # + +# A comment source is a callable that, given (kind, number), returns the raw +# Gitea comment list for that issue/PR. The route supplies a live fail-soft +# fetcher; tests supply a fixture. When None, the handoff source is reported as +# not-run (never silently empty-and-healthy). +CommentSource = Callable[[str, int], list[dict[str, Any]]] + + +@dataclass(frozen=True) +class TimelineSnapshot: + """One answered timeline query. + + ``ok`` is False when the query could not be answered as asked — currently + when a requested filter dimension no surviving source can carry was + supplied. The page is then empty *and* the snapshot says so, because an + ``ok`` empty page is a claim that no such activity exists. + """ + + schema_version: int + remote: str + org: str + repo: str + filters: dict[str, Any] + page: TimelinePage + sources: tuple[SourceStatus, ...] + ok: bool = True + error: dict[str, Any] | None = None + + def to_dict(self) -> dict[str, Any]: + return { + "ok": self.ok, + "error": self.error, + "schema_version": self.schema_version, + # Scope and filters are echoed caller input, not derived record + # data. Reflecting a value is still emitting it, so both cross the + # same boundary; ordinary scope and filter values are unchanged. + "scope": { + "remote": _safe_echo(self.remote), + "org": _safe_echo(self.org), + "repo": _safe_echo(self.repo), + }, + "filters": {key: _safe_echo(value) for key, value in self.filters.items()}, + "sources": [s.to_dict() for s in self.sources], + **self.page.to_dict(), + } + + +def _unanswerable_reasons( + statuses: Iterable[SourceStatus], unanswerable: Iterable[str] +) -> list[dict[str, str]]: + """Explain, per source, why each unanswerable dimension went unanswered.""" + out: list[dict[str, str]] = [] + for status in statuses: + for dim in unanswerable: + if dim not in status.supported_filters: + reason = _SOURCE_FILTER_LIMITS.get( + (status.name, dim), f"this source's records carry no {dim} identity" + ) + elif not status.ok: + reason = ( + f"this source can carry {dim} but did not run: " + f"{status.reason or 'unavailable'}" + ) + else: + continue + out.append({"source": status.name, "filter": dim, "reason": reason}) + return out + + +def load_timeline( + *, + remote: str, + org: str, + repo: str, + issue: int | None = None, + pr: int | None = None, + session: str | None = None, + limit: int | None = None, + offset: int | None = None, + db_path: str | None = None, + comment_source: CommentSource | None = None, +) -> TimelineSnapshot: + """Aggregate every timeline source into one filtered, paginated snapshot. + + Sources are read independently and fail soft: an unavailable source + contributes a ``SourceStatus`` with ``ok=False`` and a reason, and never + collapses the whole timeline. The handoff source only runs when a specific + issue or PR is requested (a handoff comment belongs to one thread) and a + ``comment_source`` is available; otherwise it is reported as ``not run`` + rather than as an empty-and-healthy source. + + A filter dimension that no surviving source can carry — a ``session`` + filter when the only source that ran is the control plane, whose events + record no session — is refused with ``ok=False`` and a structured error + instead of being answered with an empty page. + """ + all_events: list[WorkflowEvent] = [] + statuses: list[SourceStatus] = [] + + cp_events, cp_status = read_cp_events(remote=remote, org=org, repo=repo, db_path=db_path) + all_events.extend(cp_events) + statuses.append(cp_status) + + # Gitea handoff comments are thread-scoped: only fetch when the caller + # narrowed to one issue or PR, and only when a source was provided. + handoff_target: tuple[str, int] | None = None + if pr is not None: + handoff_target = ("pr", pr) + elif issue is not None: + handoff_target = ("issue", issue) + + if handoff_target is None: + statuses.append( + _handoff_status( + ok=False, + reason="not run: handoff comments are thread-scoped; filter by issue or pr to include them", + ) + ) + elif comment_source is None: + statuses.append( + _handoff_status( + ok=False, + reason="not run: no comment source configured for this timeline read", + ) + ) + else: + kind, number = handoff_target + try: + comments = comment_source(kind, number) or [] + handoff_events = adapt_cth_comments(comments, kind=kind, number=number) + all_events.extend(handoff_events) + statuses.append(_handoff_status(ok=True, count=len(handoff_events))) + except Exception as exc: # fail soft: a fetch/parse error degrades this source only + statuses.append(_handoff_status(ok=False, reason=f"handoff source failed: {exc}")) + + requested = tuple( + name + for name, value in ((FILTER_ISSUE, issue), (FILTER_PR, pr), (FILTER_SESSION, session)) + if value is not None + ) + statuses = [ + replace( + status, + unsupported_filters=tuple( + dim for dim in requested if dim not in status.supported_filters + ), + ) + for status in statuses + ] + filters = {"issue": issue, "pr": pr, "session": session} + + # A dimension is answerable only if a source that actually ran can carry it. + # If none can, refuse: an empty page would assert "no such activity", which + # is a claim this timeline is not in a position to make. + answerable: set[str] = set() + for status in statuses: + if status.ok: + answerable.update(status.supported_filters) + unanswerable = tuple(dim for dim in requested if dim not in answerable) + + if unanswerable: + return TimelineSnapshot( + schema_version=TIMELINE_SCHEMA_VERSION, + remote=remote, + org=org, + repo=repo, + filters=filters, + page=paginate([], limit=limit, offset=offset), + sources=tuple(statuses), + ok=False, + error={ + "code": "filter_not_supported", + "unsupported_filters": list(unanswerable), + "detail": ( + "no timeline source that ran can answer " + + ", ".join(f"'{dim}'" for dim in unanswerable) + + "; the result is refused rather than returned empty" + ), + "sources": _unanswerable_reasons(statuses, unanswerable), + }, + ) + + filtered = filter_events(all_events, issue=issue, pr=pr, session=session) + ordered = sort_events(filtered) + page = paginate(ordered, limit=limit, offset=offset) + + return TimelineSnapshot( + schema_version=TIMELINE_SCHEMA_VERSION, + remote=remote, + org=org, + repo=repo, + filters=filters, + page=page, + sources=tuple(statuses), + ) + + +def snapshot_to_dict(snapshot: TimelineSnapshot) -> dict[str, Any]: + return snapshot.to_dict() diff --git a/worktree_cleanup_audit.py b/worktree_cleanup_audit.py index 261164a..9eb4b35 100644 --- a/worktree_cleanup_audit.py +++ b/worktree_cleanup_audit.py @@ -34,7 +34,11 @@ import subprocess from datetime import datetime, timezone from typing import Any -from merged_cleanup_reconcile import branch_worktree_folder, read_local_worktree_state +from merged_cleanup_reconcile import ( + branch_worktree_folder, + is_head_ancestor_of_ref, + read_local_worktree_state, +) from reviewer_worktree import parse_dirty_tracked_files, REVIEW_WORKTREE_RE PROTECTED_BRANCHES = frozenset({"master", "main", "dev"}) @@ -67,6 +71,14 @@ REMOVABLE_CLASSES = frozenset( {CLASS_CLEAN_STALE_REMOVABLE, CLASS_DETACHED_REVIEW_LEFTOVER} ) +# Merged-PR linkage outcomes for issue worktrees (#858). Only ``LINKAGE_MERGED`` +# is ownership proof; every other outcome leaves the worktree protected. +LINKAGE_MERGED = "merged_pr" +LINKAGE_OPEN = "open_pr" +LINKAGE_NONE = "no_owning_pr" +LINKAGE_AMBIGUOUS = "ambiguous" +LINKAGE_UNKNOWN = "unknown" + _ISSUE_REF_RE = re.compile(r"issue-(\d+)", re.IGNORECASE) _ISSUE_BRANCH_PREFIXES = ("feat/", "fix/", "docs/", "chore/") @@ -169,6 +181,186 @@ def is_ttl_expired( return (now_dt - last).total_seconds() > ttl_hours * 3600.0 +def build_pr_index(prs: list[dict[str, Any]] | None) -> dict[str, list[dict[str, Any]]]: + """Index PR records by head branch for deterministic worktree linkage (#858). + + Accepts Gitea PR payloads (``head`` as a dict) and pre-flattened records + (``head_branch``/``head_sha``). Records without a usable head branch or + number are dropped rather than guessed at, so a branch is only ever linked + to a PR the caller actually proved. + """ + index: dict[str, list[dict[str, Any]]] = {} + for pr in prs or []: + head = pr.get("head") + if isinstance(head, dict): + head_branch = head.get("ref") + head_sha = head.get("sha") + else: + head_branch = pr.get("head_branch") or (head if isinstance(head, str) else None) + head_sha = pr.get("head_sha") + number = pr.get("number") + if not head_branch or number is None: + continue + try: + pr_number = int(number) + except (TypeError, ValueError): + continue + index.setdefault(str(head_branch).strip(), []).append( + { + "pr_number": pr_number, + "head_branch": str(head_branch).strip(), + "head_sha": head_sha, + "merged": bool(pr.get("merged") or pr.get("merged_at")), + "state": pr.get("state"), + } + ) + return index + + +def resolve_owning_pr( + *, + branch: str | None, + pr_index: dict[str, list[dict[str, Any]]] | None, +) -> dict[str, Any]: + """Resolve the single PR that owns ``branch``, failing closed when unclear. + + Ownership is only ``LINKAGE_MERGED`` when exactly one PR claims the branch + and that PR is merged. Several distinct PRs on one branch is a competing + claim (``LINKAGE_AMBIGUOUS``), and a still-open owner is reported as + ``LINKAGE_OPEN`` — both keep the worktree protected while still exposing + the PR number the audit resolved. + """ + if pr_index is None: + return { + "status": LINKAGE_UNKNOWN, + "pr_number": None, + "candidate_pr_numbers": [], + "reasons": ["live PR state was not supplied; ownership unproven"], + } + branch_name = (branch or "").strip() + if not branch_name: + return { + "status": LINKAGE_UNKNOWN, + "pr_number": None, + "candidate_pr_numbers": [], + "reasons": ["worktree has no attached branch; ownership unproven"], + } + + candidates = list(pr_index.get(branch_name) or []) + numbers = sorted({c["pr_number"] for c in candidates}) + if not candidates: + return { + "status": LINKAGE_NONE, + "pr_number": None, + "candidate_pr_numbers": [], + "reasons": [f"no PR claims branch '{branch_name}'"], + } + if len(numbers) > 1: + return { + "status": LINKAGE_AMBIGUOUS, + "pr_number": None, + "candidate_pr_numbers": numbers, + "reasons": [ + f"branch '{branch_name}' is claimed by competing PRs {numbers}; " + "ownership is ambiguous" + ], + } + + owner = candidates[0] + pr_number = owner["pr_number"] + if owner.get("head_branch") != branch_name: + return { + "status": LINKAGE_UNKNOWN, + "pr_number": pr_number, + "candidate_pr_numbers": numbers, + "reasons": [ + f"PR #{pr_number} head branch '{owner.get('head_branch')}' does not " + f"match worktree branch '{branch_name}'" + ], + } + if not owner.get("merged"): + return { + "status": LINKAGE_OPEN, + "pr_number": pr_number, + "candidate_pr_numbers": numbers, + "pr_head_sha": owner.get("head_sha"), + "reasons": [f"owning PR #{pr_number} is not merged"], + } + return { + "status": LINKAGE_MERGED, + "pr_number": pr_number, + "candidate_pr_numbers": numbers, + "pr_head_sha": owner.get("head_sha"), + "reasons": [], + } + + +def assess_merged_pr_worktree_cleanup( + *, + linkage: dict[str, Any] | None, + head_sha: str | None, + head_in_master: bool | None, + is_dirty: bool, + has_open_pr: bool, + has_active_lease: bool, + has_active_issue_lock: bool, + is_protected: bool, + has_live_session: bool = False, +) -> dict[str, Any]: + """Decide whether a merged issue worktree satisfies the full cleanup policy. + + Every condition must be independently proven: conclusive merged-PR + ownership, agreement between the worktree branch and the PR head branch, + containment of the worktree head in authoritative master (which is what + proves no unmerged commits remain), absence of any open/competing PR, + lease, issue lock, or live session, a clean tree, and a worktree that is + not the protected control checkout. Anything unknown blocks. + """ + link = linkage or { + "status": LINKAGE_UNKNOWN, + "pr_number": None, + "reasons": ["no linkage assessment supplied"], + } + status = link.get("status") + reasons: list[str] = [] + + if status != LINKAGE_MERGED: + reasons.extend( + link.get("reasons") or ["owning PR could not be conclusively identified"] + ) + if is_protected: + reasons.append("worktree is protected or the stable control checkout") + if is_dirty: + reasons.append("worktree has uncommitted changes") + if has_open_pr: + reasons.append("worktree branch has an open PR") + if has_active_lease: + reasons.append("worktree has an active lease") + if has_active_issue_lock: + reasons.append("an active issue lock references this branch") + if has_live_session: + reasons.append("a live process or session is using this worktree") + if not head_sha: + reasons.append("worktree head sha is unknown") + if head_in_master is None: + reasons.append("containment of the worktree head in master is unknown") + elif not head_in_master: + reasons.append( + "worktree head is not contained in authoritative master " + "(unmerged commits remain)" + ) + + proven = not reasons + return { + "linkage_status": status, + "pr_number": link.get("pr_number"), + "pr_head_sha": link.get("pr_head_sha"), + "head_in_master": head_in_master, + "proven": proven, + "block_reasons": reasons, + } + + def classify_worktree( *, workflow_type: str, @@ -181,6 +373,8 @@ def classify_worktree( ttl_expired: bool = False, is_protected: bool = False, metadata_known: bool = True, + merged_pr_cleanup: dict[str, Any] | None = None, + has_live_session: bool = False, ) -> str: """Classify a worktree, safety-first: any preservation signal wins. @@ -199,6 +393,8 @@ def classify_worktree( return CLASS_ACTIVE_ISSUE_WORK # never auto-deleted (criterion 8) if has_active_issue_lock: return CLASS_ACTIVE_ISSUE_WORK + if has_live_session: + return CLASS_ACTIVE_ISSUE_WORK # a live session still owns this tree if not metadata_known or workflow_type == WORKFLOW_UNKNOWN: return CLASS_UNSAFE_UNKNOWN # never auto-deleted without proof @@ -207,7 +403,15 @@ def classify_worktree( if is_detached or branch_gone: return CLASS_DETACHED_REVIEW_LEFTOVER return CLASS_CLEAN_STALE_REMOVABLE - # issue_work / conflict_fix: only removable once the TTL has expired. + if workflow_type == WORKFLOW_ISSUE_WORK: + # #858: an issue worktree becomes removable only on authoritative + # merged-PR evidence satisfying the whole cleanup policy. Age alone + # never proves the branch landed, so TTL cannot qualify one by itself + # — otherwise a worktree holding unmerged commits would be reclaimed. + if (merged_pr_cleanup or {}).get("proven"): + return CLASS_CLEAN_STALE_REMOVABLE + return CLASS_ACTIVE_ISSUE_WORK + # conflict_fix: only removable once the TTL has expired. if ttl_expired: return CLASS_CLEAN_STALE_REMOVABLE return CLASS_ACTIVE_ISSUE_WORK @@ -400,6 +604,20 @@ def remove_worktree(project_root: str, path: str) -> dict[str, Any]: } +def head_contained_in_ref( + project_root: str, head_sha: str | None, ref: str | None +) -> bool | None: + """Return True when ``head_sha`` is already contained in ``ref``. + + Shares :mod:`merged_cleanup_reconcile`'s ancestry check so the audit and + the PR-scoped reconciler agree on what "already landed" means (#858). + Returns None when containment cannot be determined, which fails closed. + """ + if not head_sha or not ref: + return None + return is_head_ancestor_of_ref(project_root, head_sha, ref) + + def _is_under_branches(project_root: str, path: str) -> bool: branches_root = os.path.join(os.path.abspath(project_root), "branches") return os.path.abspath(path or "").startswith(branches_root + os.sep) @@ -413,16 +631,30 @@ def audit_branches_directory( active_issue_branches: set[str] | None = None, now: datetime | str | None = None, ttl_hours: float = DEFAULT_TTL_HOURS, + pr_index: dict[str, list[dict[str, Any]]] | None = None, + leased_issue_numbers: set[int] | None = None, + live_session_paths: set[str] | None = None, + master_ref: str | None = None, ) -> dict[str, Any]: """Classify every session-owned worktree under ``branches/``. Read-only: shells out to git for discovery and dirty state, then applies the pure classifier. Returns per-worktree classifications, counts, the list of removable candidates, and the ``git worktree list`` proof. + + ``pr_index`` (see :func:`build_pr_index`) supplies the authoritative PR + ownership used to link issue worktrees to their merged PR (#858). + ``master_ref`` is the ref a worktree head must be contained in before it + can be considered landed. Both are optional and their absence only ever + fails closed: without them no issue worktree becomes removable. """ open_pr_branches = open_pr_branches or set() leased_branches = leased_branches or set() active_issue_branches = active_issue_branches or set() + leased_issue_numbers = leased_issue_numbers or set() + live_session_paths = { + os.path.abspath(p) for p in (live_session_paths or set()) if p + } worktrees: list[dict[str, Any]] = [] for entry in list_worktrees(project_root): @@ -433,12 +665,42 @@ def audit_branches_directory( ) dirty_state = read_worktree_dirty(path) is_dirty = bool(dirty_state.get("dirty")) + head_sha = entry.get("head") + linkage = resolve_owning_pr(branch=branch, pr_index=pr_index) metadata = build_worktree_metadata( - path=path, branch=branch, head_sha=entry.get("head") + path=path, + branch=branch, + head_sha=head_sha, + pr_number=linkage.get("pr_number"), ) has_open_pr = bool(branch) and branch in open_pr_branches - has_active_lease = bool(branch) and branch in leased_branches + # A lease on issue N protects that issue's own work worktree. It must + # not incidentally protect a baseline/review scratch tree that merely + # carries the same issue marker in its name, which would change the + # classification of worktrees this policy does not own. + has_active_lease = (bool(branch) and branch in leased_branches) or ( + metadata["workflow_type"] == WORKFLOW_ISSUE_WORK + and metadata.get("issue_number") is not None + and metadata["issue_number"] in leased_issue_numbers + ) has_active_lock = bool(branch) and branch in active_issue_branches + has_live_session = bool(path) and os.path.abspath(path) in live_session_paths + head_in_master = ( + head_contained_in_ref(project_root, head_sha, master_ref) + if master_ref + else None + ) + merged_pr_cleanup = assess_merged_pr_worktree_cleanup( + linkage=linkage, + head_sha=head_sha, + head_in_master=head_in_master, + is_dirty=is_dirty, + has_open_pr=has_open_pr, + has_active_lease=has_active_lease, + has_active_issue_lock=has_active_lock, + is_protected=is_protected, + has_live_session=has_live_session, + ) ttl_expired = is_ttl_expired( last_used_at=metadata.get("last_used_at"), now=now, ttl_hours=ttl_hours ) @@ -452,6 +714,8 @@ def audit_branches_directory( branch_gone=branch is None and not entry.get("detached"), ttl_expired=ttl_expired, is_protected=is_protected, + merged_pr_cleanup=merged_pr_cleanup, + has_live_session=has_live_session, ) metadata["cleanup_eligibility"] = classification worktrees.append( @@ -463,7 +727,10 @@ def audit_branches_directory( "has_open_pr": has_open_pr, "has_active_lease": has_active_lease, "has_active_issue_lock": has_active_lock, + "has_live_session": has_live_session, "is_protected": is_protected, + "merged_pr_linkage": linkage, + "merged_pr_cleanup": merged_pr_cleanup, "classification": classification, "removable": is_removable(classification), }