fix(mcp): forward worktree_path into publication preflight (Closes #815)

gitea_publish_unpublished_issue_branch accepted a required worktree_path
but resolved it only after verify_preflight_purity had already run, so the
#618 branches-only guard and every other workspace-resolution layer behind
preflight received None and fell back to the MCP process root. A daemon
rooted at the stable control checkout therefore refused a valid registered
issue worktree the caller had explicitly supplied, before the publication
assessor could use it — the sole verify_preflight_purity call site that
accepted a worktree argument and dropped it.

Resolve the workspace once, before preflight, and forward it. A blank or
absent path forwards None and keeps the ordinary #618 fail-closed fallback,
so guard strictness is unchanged for missing, empty, unregistered, foreign,
or control-checkout worktrees. Public tool contract, ownership, cleanliness,
hash, ancestry, and read-after-write protections from PR #814 are untouched.

Adds tests/test_issue_815_preflight_worktree_forwarding.py: a #735-style
capture proving the argument reaches verify_preflight_purity, a faithful
production reproduction (control-rooted daemon, no session lock, explicit
worktree) that clears #618 on the fixed source and is trapped at #618 on the
unpatched source, an end-to-end control-rooted publication in the real
topology (PROJECT_ROOT is the stable control checkout, the issue worktree is
a distinct registered path, production guards forced on), and negative
coverage keeping every #618 and assessor refusal intact. The prior #812
suite masked the defect by patching PROJECT_ROOT to equal the issue worktree.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>

anti_stomp_preflight.py

@@ -87,6 +87,7 @@ MUTATION_TASKS = frozenset({
     "edit_pr",
     "commit_files",
     "gitea_commit_files",
+    "publish_unpublished_branch",
     "delete_branch",
     "cleanup_merged_pr_branch",
     "cleanup_stale_claims",

branch_publish.py

@@ -0,0 +1,591 @@
+"""Publish an unpublished local commit on a registered issue worktree (#812 AC20).
+
+Entry point B of #812 is the state where an author's work has already advanced
+to a local commit: the worktree is registered, clean, on the issue branch, and
+carries the only copy of the implementation, but the branch has never been
+published. That state deadlocks, because two individually correct predicates
+close a cycle:
+
+* ``issue_lock_renewal.assess_exact_owner_lease_renewal`` refuses to renew an
+  expired lease without an observable remote head — an unpublished branch has
+  none.
+* Every publication path (``gitea_commit_files``, ``gitea_create_pr``) derives
+  its workspace from the author issue lock under #618, so nothing can create
+  that remote head without first holding the lock.
+
+This module supplies the missing operation: it publishes an *already committed*
+local head to the remote branch, so exact-owner renewal has the evidence it
+requires. It deliberately does **not** renew, reclaim, rebind, or clear any
+lock. Publication is the whole of its authority.
+
+Why this is not a lock bypass
+-----------------------------
+The operation can only publish a branch whose **durable issue-lock record
+already names the caller as claimant**. Ownership is read from the lock file on
+disk (``issue_lock_store``), never from a caller-supplied flag, so the tool
+cannot manufacture a claim it does not already hold. Nothing here weakens the
+#510/#618/#713 guards: a dirty tree, an unregistered worktree, a foreign
+claimant, a changed HEAD, or a divergent remote head each refuse, exactly as
+they do today. The only thing this adds is the ability to make an existing,
+owned, committed, clean branch observable on the remote.
+
+Separation of records (#812 AC23)
+---------------------------------
+The durable **issue-lock file** and the control-plane **workflow lease** are
+distinct records. This module reads the former as ownership evidence and writes
+neither. Publishing changes remote git state only; no lock is renewed,
+abandoned, reclaimed, or generation-bumped here.
+
+Process evidence (#812 AC24)
+----------------------------
+Liveness of the lock's recorded pid is **not consulted**. That is deliberate:
+the recorded pid routinely belongs to the long-running MCP daemon rather than to
+an active author client, and the existing reclaim predicate
+(``assess_expired_lock_reclaim``) can never be satisfied while that daemon runs.
+Publication does not require the recording process to be dead, so this module
+never asserts, infers, or depends on a process being dead. Ownership is proven
+by identity and profile match against the recorded claimant instead.
+"""
+
+from __future__ import annotations
+
+import hashlib
+import os
+import re
+import subprocess
+
+from reviewer_worktree import parse_dirty_tracked_files
+from stable_branch_push_guard import is_stable_ref, redact_command
+
+# Assessment outcomes.
+PUBLISH_SANCTIONED = "publish_sanctioned"
+ALREADY_PUBLISHED = "already_published"
+REFUSED = "refused"
+
+#: Implementation branches must stay traceable to their issue (#713 lineage).
+ISSUE_BRANCH_RE = re.compile(r"^(fix|feat|docs|chore)/issue-(\d+)-.+$")
+
+_SHA_RE = re.compile(r"^[0-9a-f]{40}$")
+
+
+def _text(value: object) -> str:
+    return value.strip() if isinstance(value, str) else ""
+
+
+def _realpath(value: str | None) -> str | None:
+    path = _text(value)
+    return os.path.realpath(path) if path else None
+
+
+def parse_untracked_files(porcelain_status: str) -> list[str]:
+    """Return untracked paths from ``git status --porcelain`` output.
+
+    ``reviewer_worktree.parse_dirty_tracked_files`` deliberately skips ``??``
+    entries. Publication needs both halves: an untracked file in the worktree is
+    unpublished content that the commit does not carry, so publishing would
+    silently leave it behind.
+    """
+    untracked: list[str] = []
+    for line in (porcelain_status or "").splitlines():
+        if not line.startswith("??"):
+            continue
+        path = line[2:].strip()
+        if path:
+            untracked.append(path)
+    return untracked
+
+
+def hash_worktree_files(worktree_path: str, paths) -> dict[str, str | None]:
+    """SHA-256 each path under *worktree_path*; ``None`` when unreadable."""
+    root = _text(worktree_path)
+    hashes: dict[str, str | None] = {}
+    for rel in paths or ():
+        rel_text = _text(rel)
+        if not rel_text:
+            continue
+        full = os.path.join(root, rel_text)
+        try:
+            with open(full, "rb") as handle:
+                digest = hashlib.sha256()
+                for chunk in iter(lambda: handle.read(65536), b""):
+                    digest.update(chunk)
+            hashes[rel_text] = digest.hexdigest()
+        except OSError:
+            hashes[rel_text] = None
+    return hashes
+
+
+def read_remote_branch_head(
+    worktree_path: str, remote_name: str, branch_name: str
+) -> dict:
+    """Observe the remote head for *branch_name*, read-only.
+
+    ``probe_ok`` False means git could not answer at all. That is kept distinct
+    from "the branch does not exist": an unobservable remote must fail closed
+    rather than be mistaken for an absent branch, because the two lead to
+    opposite dispositions.
+    """
+    path = _text(worktree_path)
+    remote = _text(remote_name)
+    branch = _text(branch_name)
+    result: dict = {
+        "probe_ok": False,
+        "remote_branch_exists": False,
+        "remote_head_sha": None,
+        "reasons": [],
+    }
+    if not (path and remote and branch):
+        result["reasons"].append(
+            "remote head probe requires a worktree path, remote name, and branch"
+        )
+        return result
+    try:
+        res = subprocess.run(
+            ["git", "-C", path, "ls-remote", remote, f"refs/heads/{branch}"],
+            capture_output=True,
+            text=True,
+            check=False,
+        )
+    except OSError as exc:  # git unavailable — fail closed, never assume absent
+        result["reasons"].append(f"remote head probe could not run: {exc}")
+        return result
+    if res.returncode != 0:
+        result["reasons"].append(
+            f"remote head probe failed for '{branch}' on remote '{remote}'"
+        )
+        return result
+
+    result["probe_ok"] = True
+    for line in (res.stdout or "").splitlines():
+        parts = line.split()
+        if len(parts) >= 2 and parts[1] == f"refs/heads/{branch}":
+            result["remote_branch_exists"] = True
+            result["remote_head_sha"] = parts[0].strip()
+            break
+    return result
+
+
+def read_is_ancestor(
+    worktree_path: str, ancestor_sha: str, descendant_sha: str
+) -> dict:
+    """Observe whether *ancestor_sha* is an ancestor of *descendant_sha*."""
+    path = _text(worktree_path)
+    ancestor = _text(ancestor_sha)
+    descendant = _text(descendant_sha)
+    result: dict = {"probe_ok": False, "is_ancestor": False, "reasons": []}
+    if not (path and ancestor and descendant):
+        result["reasons"].append(
+            "ancestry probe requires a worktree path and both commit SHAs"
+        )
+        return result
+    try:
+        present = subprocess.run(
+            ["git", "-C", path, "rev-parse", "--verify", "--quiet",
+             f"{ancestor}^{{commit}}"],
+            capture_output=True, text=True, check=False,
+        )
+        if present.returncode != 0:
+            result["reasons"].append(
+                f"remote head {ancestor} is not present locally, so it cannot be "
+                "proven an ancestor of the commit being published"
+            )
+            return result
+        res = subprocess.run(
+            ["git", "-C", path, "merge-base", "--is-ancestor", ancestor, descendant],
+            capture_output=True, text=True, check=False,
+        )
+    except OSError as exc:
+        result["reasons"].append(f"ancestry probe could not run: {exc}")
+        return result
+    result["probe_ok"] = res.returncode in (0, 1)
+    result["is_ancestor"] = res.returncode == 0
+    return result
+
+
+def assess_unpublished_commit_publication(
+    existing_lock,
+    *,
+    issue_number: int,
+    branch_name: str,
+    worktree_path: str,
+    expected_head: str,
+    remote: str,
+    org: str,
+    repo: str,
+    identity: str | None,
+    profile: str | None,
+    worktree_state,
+    worktree_registered: bool | None = None,
+    remote_probe=None,
+    ancestry=None,
+    competing_open_prs=(),
+    expected_file_hashes=None,
+    observed_file_hashes=None,
+) -> dict:
+    """Decide whether an unpublished local commit may be published (#812 AC20).
+
+    Pure predicate. Every input is either a caller-declared expectation that
+    must be *matched* against observation, or a server-side observation. No
+    caller-supplied boolean is accepted as proof of ownership, liveness, or
+    eligibility: ``existing_lock`` comes from the durable lock file and the
+    git/PR state is observed by the server.
+
+    The single mutating disposition it can return is "publish this exact commit
+    to this exact branch". It never sanctions renewal, reclamation, force
+    updates, history rewriting, or publication of uncommitted content.
+    """
+    reasons: list[str] = []
+    branch = _text(branch_name)
+    head = _text(expected_head).lower()
+    workspace = _realpath(worktree_path)
+    state = worktree_state if isinstance(worktree_state, dict) else {}
+    lock = existing_lock if isinstance(existing_lock, dict) else None
+
+    evidence: dict = {
+        "issue_number": issue_number,
+        "branch_name": branch or None,
+        "worktree_path": workspace,
+        "expected_head": head or None,
+        "remote": _text(remote) or None,
+        "org": _text(org) or None,
+        "repo": _text(repo) or None,
+        "identity": _text(identity) or None,
+        "profile": _text(profile) or None,
+        "lock_record_present": lock is not None,
+        "recorded_claimant": None,
+        "recorded_branch": None,
+        "recorded_worktree": None,
+        "lock_generation": None,
+        "local_head_sha": _text(state.get("head_sha")) or None,
+        "current_branch": _text(state.get("current_branch")) or None,
+        "dirty_tracked_files": [],
+        "untracked_files": [],
+        "worktree_registered": worktree_registered,
+        "remote_branch_exists": None,
+        "remote_head_sha": None,
+        "fast_forward_from_remote": None,
+        "competing_open_prs": [],
+        "file_hashes_verified": None,
+        "hash_mismatches": [],
+        # Recorded explicitly so no reader mistakes silence for a liveness
+        # claim, and so the audit shows which records were left alone (AC23/AC24).
+        "owner_pid_liveness_consulted": False,
+        "workflow_lease_touched": False,
+        "issue_lock_record_mutated": False,
+    }
+
+    # ── declared shape ────────────────────────────────────────────────────
+    if not branch:
+        reasons.append("branch name not declared; fail closed")
+    if not head:
+        reasons.append(
+            "expected_head not declared; publication must name the exact commit"
+        )
+    elif not _SHA_RE.match(head):
+        reasons.append(
+            f"expected_head '{head}' is not a full 40-character commit SHA; "
+            "abbreviated or symbolic revisions are refused"
+        )
+    if not workspace:
+        reasons.append("worktree path not declared; fail closed")
+
+    if branch:
+        match = ISSUE_BRANCH_RE.match(branch)
+        if not match:
+            reasons.append(
+                f"branch '{branch}' is not an issue-linked implementation branch "
+                "((fix|feat|docs|chore)/issue-<number>-<description>); fail closed"
+            )
+        elif int(match.group(2)) != int(issue_number):
+            reasons.append(
+                f"branch '{branch}' does not carry issue number {issue_number}; "
+                "fail closed"
+            )
+        if is_stable_ref(branch):
+            reasons.append(
+                f"refusing to publish stable branch '{branch}'; this operation "
+                "publishes issue branches only"
+            )
+
+    # ── ownership: durable issue-lock record only (AC8, AC20, AC24) ───────
+    if lock is None:
+        reasons.append(
+            "no durable issue-lock record for this issue; publication requires an "
+            "existing recorded claim naming the caller, so this operation cannot "
+            "be used to bypass the author lock"
+        )
+    else:
+        lease = lock.get("work_lease")
+        lease = lease if isinstance(lease, dict) else {}
+        claimant = lease.get("claimant")
+        claimant = claimant if isinstance(claimant, dict) else {}
+        recorded_user = _text(claimant.get("username"))
+        recorded_profile = _text(claimant.get("profile"))
+        recorded_branch = _text(lock.get("branch_name")) or _text(lease.get("branch"))
+        recorded_worktree = _realpath(
+            _text(lock.get("worktree_path")) or _text(lease.get("worktree_path"))
+        )
+        evidence["recorded_claimant"] = {
+            "username": recorded_user or None,
+            "profile": recorded_profile or None,
+        }
+        evidence["recorded_branch"] = recorded_branch or None
+        evidence["recorded_worktree"] = recorded_worktree
+        try:
+            evidence["lock_generation"] = int(lock.get("lock_generation") or 0)
+        except (TypeError, ValueError):
+            evidence["lock_generation"] = 0
+
+        try:
+            recorded_issue = int(lock.get("issue_number") or 0)
+        except (TypeError, ValueError):
+            recorded_issue = 0
+        if recorded_issue != int(issue_number):
+            reasons.append(
+                f"durable lock records issue {lock.get('issue_number')}, not "
+                f"{issue_number}; ambiguous ownership, fail closed"
+            )
+        for field, declared in (
+            ("remote", _text(remote)),
+            ("org", _text(org)),
+            ("repo", _text(repo)),
+        ):
+            recorded = _text(lock.get(field))
+            if recorded and declared and recorded != declared:
+                reasons.append(
+                    f"durable lock records {field} '{recorded}' but the request "
+                    f"declares '{declared}'; repository mismatch, fail closed"
+                )
+        if recorded_branch and branch and recorded_branch != branch:
+            reasons.append(
+                f"durable lock records branch '{recorded_branch}' but the request "
+                f"declares '{branch}'; fail closed"
+            )
+        if recorded_worktree and workspace and recorded_worktree != workspace:
+            reasons.append(
+                f"durable lock records worktree '{recorded_worktree}' but the "
+                f"request declares '{workspace}'; fail closed"
+            )
+        if not recorded_user or not recorded_profile:
+            reasons.append(
+                "durable lock does not record a claimant username and profile; "
+                "ownership cannot be proven, fail closed"
+            )
+        else:
+            if recorded_user != _text(identity):
+                reasons.append(
+                    f"durable lock claimant '{recorded_user}' is not the acting "
+                    f"identity '{_text(identity) or '(unknown)'}'; foreign claim, "
+                    "fail closed"
+                )
+            if recorded_profile != _text(profile):
+                reasons.append(
+                    f"durable lock claimant profile '{recorded_profile}' is not "
+                    f"the active profile '{_text(profile) or '(unknown)'}'; "
+                    "fail closed"
+                )
+
+    # ── worktree: registered, on-branch, clean, at the expected commit ────
+    if worktree_registered is False:
+        reasons.append(
+            f"worktree '{workspace}' is not listed in git worktree list; #713 "
+            "requires a genuinely registered worktree, fail closed"
+        )
+
+    current_branch = _text(state.get("current_branch"))
+    if not current_branch:
+        reasons.append("worktree branch could not be observed; fail closed")
+    elif branch and current_branch != branch:
+        reasons.append(
+            f"worktree is on branch '{current_branch}', not '{branch}'; fail closed"
+        )
+
+    porcelain = state.get("porcelain_status") or ""
+    dirty_tracked = parse_dirty_tracked_files(porcelain)
+    untracked = parse_untracked_files(porcelain)
+    evidence["dirty_tracked_files"] = dirty_tracked
+    evidence["untracked_files"] = untracked
+    if dirty_tracked:
+        reasons.append(
+            "worktree has dirty tracked files, so the commit is not the whole of "
+            f"the work: {', '.join(dirty_tracked)}. This operation publishes an "
+            "existing clean commit only; uncommitted content is out of scope"
+        )
+    if untracked:
+        reasons.append(
+            "worktree has untracked files that the commit does not carry: "
+            f"{', '.join(untracked)}. Publishing would silently leave them "
+            "behind; fail closed"
+        )
+
+    local_head = _text(state.get("head_sha")).lower()
+    if not local_head:
+        reasons.append("local HEAD could not be observed; fail closed")
+    elif head and local_head != head:
+        reasons.append(
+            f"worktree HEAD is {local_head} but the request declares {head}; the "
+            "local commit changed since it was recorded, fail closed"
+        )
+
+    # ── remote state ──────────────────────────────────────────────────────
+    probe = remote_probe if isinstance(remote_probe, dict) else {}
+    already_published = False
+    if not probe.get("probe_ok"):
+        reasons.append(
+            "remote branch head could not be observed; publication must not "
+            "proceed against an unknown remote state, fail closed"
+        )
+        reasons.extend(probe.get("reasons") or [])
+    else:
+        remote_exists = bool(probe.get("remote_branch_exists"))
+        remote_head = _text(probe.get("remote_head_sha")).lower() or None
+        evidence["remote_branch_exists"] = remote_exists
+        evidence["remote_head_sha"] = remote_head
+        if remote_exists and remote_head and head:
+            if remote_head == head:
+                already_published = True
+                evidence["fast_forward_from_remote"] = True
+            else:
+                anc = ancestry if isinstance(ancestry, dict) else {}
+                is_anc = bool(anc.get("probe_ok")) and bool(anc.get("is_ancestor"))
+                evidence["fast_forward_from_remote"] = is_anc
+                if not is_anc:
+                    reasons.append(
+                        f"remote branch '{branch}' already exists at {remote_head}, "
+                        f"which is not an ancestor of {head}; publishing would "
+                        "discard or rewrite published history, fail closed"
+                    )
+                    reasons.extend(anc.get("reasons") or [])
+        elif remote_exists and not remote_head:
+            reasons.append(
+                f"remote branch '{branch}' exists but its head could not be read; "
+                "fail closed"
+            )
+
+    # ── competing claims ──────────────────────────────────────────────────
+    competing = [p for p in (competing_open_prs or ()) if p]
+    evidence["competing_open_prs"] = list(competing)
+    if competing:
+        reasons.append(
+            f"open pull request(s) {competing} already claim issue {issue_number} "
+            "or this branch; ambiguous ownership, fail closed"
+        )
+
+    # ── content verification before publication ───────────────────────────
+    if expected_file_hashes:
+        observed = (
+            observed_file_hashes if isinstance(observed_file_hashes, dict) else {}
+        )
+        mismatches: list[str] = []
+        for path, expected_digest in dict(expected_file_hashes).items():
+            actual = observed.get(path)
+            if actual is None:
+                mismatches.append(f"{path}: missing or unreadable in the worktree")
+            elif _text(actual).lower() != _text(expected_digest).lower():
+                mismatches.append(
+                    f"{path}: expected {expected_digest}, observed {actual}"
+                )
+        evidence["hash_mismatches"] = mismatches
+        evidence["file_hashes_verified"] = not mismatches
+        if mismatches:
+            reasons.append(
+                "declared content hashes do not match the worktree: "
+                + "; ".join(mismatches)
+                + ". Refusing to publish content that is not what was recorded"
+            )
+
+    if reasons:
+        return {
+            "outcome": REFUSED,
+            "publish_sanctioned": False,
+            "already_published": False,
+            "reasons": reasons,
+            "evidence": evidence,
+        }
+    return {
+        "outcome": ALREADY_PUBLISHED if already_published else PUBLISH_SANCTIONED,
+        # Idempotent retry: a remote head that already equals the assessed commit
+        # needs no second push, so the caller verifies instead of acting.
+        "publish_sanctioned": not already_published,
+        "already_published": already_published,
+        "reasons": [],
+        "evidence": evidence,
+    }
+
+
+def publish_commit_to_remote_branch(
+    *,
+    worktree_path: str,
+    remote_name: str,
+    branch_name: str,
+    expected_head: str,
+) -> dict:
+    """Send exactly *expected_head* to ``refs/heads/<branch_name>``.
+
+    The refspec names the commit SHA explicitly rather than ``HEAD`` or the
+    local branch, so what lands is the commit that was assessed and nothing
+    else. No force, no lease, no ``+`` prefix: a non-fast-forward is rejected by
+    git itself, the last of several independent guards against overwriting
+    published history.
+    """
+    path = _text(worktree_path)
+    remote = _text(remote_name)
+    branch = _text(branch_name)
+    head = _text(expected_head)
+    result: dict = {
+        "success": False,
+        "pushed_ref": f"refs/heads/{branch}" if branch else None,
+        "pushed_sha": head or None,
+        "stderr": None,
+        "reasons": [],
+    }
+    if not (path and remote and branch and head):
+        result["reasons"].append(
+            "publication requires a worktree path, remote, branch, and commit SHA"
+        )
+        return result
+
+    refspec = f"{head}:refs/heads/{branch}"
+    try:
+        res = subprocess.run(
+            ["git", "-C", path, "push", remote, refspec],
+            capture_output=True,
+            text=True,
+            check=False,
+        )
+    except OSError as exc:
+        result["reasons"].append(f"publication could not run: {exc}")
+        return result
+
+    if res.returncode != 0:
+        # Redact before surfacing: failures can echo credentialed remote URLs.
+        result["stderr"] = redact_command(res.stderr or "")
+        result["reasons"].append(
+            f"publication of {head} to '{branch}' on remote '{remote}' failed"
+        )
+        return result
+
+    result["success"] = True
+    return result
+
+
+def verify_published_head(
+    *, worktree_path: str, remote_name: str, branch_name: str, expected_head: str
+) -> dict:
+    """Read-after-write: confirm the remote head equals *expected_head* (AC20)."""
+    probe = read_remote_branch_head(worktree_path, remote_name, branch_name)
+    head = _text(expected_head).lower()
+    observed = _text(probe.get("remote_head_sha")).lower() or None
+    verified = bool(head) and bool(probe.get("probe_ok")) and observed == head
+    reasons: list[str] = list(probe.get("reasons") or [])
+    if probe.get("probe_ok") and not verified:
+        reasons.append(
+            "read-after-write verification failed: remote head is "
+            f"{observed or '(absent)'}, expected {head}"
+        )
+    return {
+        "verified": verified,
+        "remote_head_sha": observed,
+        "expected_head": head or None,
+        "reasons": reasons,
+    }

docs/mcp-tool-inventory.md

@@ -120,6 +120,7 @@ that gates each call, not which tools exist.
 - `gitea_observability_list_projects`
 - `gitea_observability_reconcile_incident`
 - `gitea_post_heartbeat`
+- `gitea_publish_unpublished_issue_branch`
 - `gitea_quarantine_contaminated_review`
 - `gitea_reclaim_expired_workflow_lease`
 - `gitea_reconcile_already_landed_pr`

gitea_mcp_server.py

@@ -2024,6 +2024,7 @@ import review_quarantine  # noqa: E402  # #695 contaminated formal-review quaran
 import mcp_daemon_guard  # noqa: E402  # #695 native transport provenance
 import already_landed_reconcile  # noqa: E402
 import author_mutation_worktree  # noqa: E402
+import branch_publish  # noqa: E402  # #812 AC20 unpublished-commit publication
 import root_checkout_guard  # noqa: E402
 import workflow_scope_guard  # noqa: E402  # #683 production scope / force-on guards
 import stable_branch_push_guard  # noqa: E402
@@ -9082,6 +9083,253 @@ def gitea_commit_files(
     }
 
 
+def _publication_block(reasons: list[str], **extra) -> dict:
+    """Uniform fail-closed shape for publication refusals (#812 AC20)."""
+    payload = {
+        "success": False,
+        "performed": False,
+        "published": False,
+        "verified": False,
+        "outcome": branch_publish.REFUSED,
+        "reasons": reasons,
+        # State every record this operation left alone, so a refusal can never
+        # be misread as a lock mutation (#812 AC23).
+        "issue_lock_record_mutated": False,
+        "workflow_lease_touched": False,
+    }
+    payload.update(extra)
+    return payload
+
+
+@mcp.tool()
+def gitea_publish_unpublished_issue_branch(
+    issue_number: int,
+    branch_name: str,
+    worktree_path: str,
+    expected_head: str,
+    remote: str = "dadeschools",
+    host: str | None = None,
+    org: str | None = None,
+    repo: str | None = None,
+    git_remote_name: str | None = None,
+    expected_file_hashes: dict | None = None,
+    dry_run: bool = False,
+) -> dict:
+    """Publish an already-committed, unpublished issue branch (#812 AC20).
+
+    Creates the remote head for a branch whose work is *already* a local commit
+    on a registered, clean worktree, so exact-owner lease renewal
+    (``issue_lock_renewal``) has the published head its evidence model requires.
+    This is the one step of the entry point B deadlock that no existing tool can
+    perform: publication is otherwise lock-derived under #618, and the lock
+    itself is withheld until a remote head exists.
+
+    Not a lock bypass. The branch's **durable issue-lock record must already
+    name the caller as claimant** — ownership is read from the lock file, never
+    asserted by the caller — so this can only publish work the caller already
+    owns. It refuses a dirty or untracked-carrying worktree, an unregistered
+    worktree, a changed local HEAD, a remote head that is not an ancestor of the
+    commit, a competing open PR on another branch for the same issue, and any
+    declared-hash mismatch. It renews, reclaims, and clears nothing: the durable
+    issue-lock file and the control-plane workflow lease are both left untouched
+    (#812 AC23), and the recorded owner pid's liveness is never consulted or
+    asserted (#812 AC24).
+
+    Args:
+        issue_number: The issue whose recorded claim authorizes publication.
+        branch_name: Issue branch to publish, ``(fix|feat|docs|chore)/issue-N-…``.
+        worktree_path: Registered worktree holding the commit.
+        expected_head: Full 40-character SHA of the commit to publish. Required:
+            publication names the exact commit, and a mismatch fails closed.
+        remote: Known instance — 'dadeschools' or 'prgs'.
+        host: Override the Gitea host.
+        org: Override the owner/organization.
+        repo: Override the repository name.
+        git_remote_name: Git remote to publish to; defaults to *remote*.
+        expected_file_hashes: Optional ``{path: sha256}`` verified against the
+            worktree before publication. Any mismatch or missing file refuses.
+        dry_run: Report the decision and evidence, mutate nothing.
+
+    Returns:
+        dict with 'success', 'performed', 'published', 'verified', 'outcome',
+        'remote_head_sha', 'reasons', and 'evidence'.
+    """
+    task = "publish_unpublished_branch"
+    ok, block_reasons = role_session_router.check_author_mutation_after_reviewer_stop(
+        task
+    )
+    if not ok:
+        return _publication_block(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
+
+    # #815: resolve the caller's worktree *before* preflight and forward it, so
+    # every workspace-resolution layer behind verify_preflight_purity — including
+    # the #618 branches-only guard — judges the registered issue worktree this
+    # publication actually operates on. Resolving it afterwards let preflight
+    # fall back to the MCP process root, so a daemon rooted at the stable control
+    # checkout refused a valid explicit worktree before the assessor ever ran.
+    # A caller supplying nothing usable forwards None and keeps the ordinary
+    # fail-closed fallback.
+    explicit_worktree = (worktree_path or "").strip()
+    workspace = os.path.realpath(os.path.abspath(explicit_worktree or "."))
+
+    verify_preflight_purity(
+        remote,
+        worktree_path=workspace if explicit_worktree else None,
+        task=task,
+        org=org,
+        repo=repo,
+    )
+
+    h, o, r = _resolve(remote, host, org, repo)
+    git_remote = (git_remote_name or remote or "").strip()
+
+    existing_lock = issue_lock_store.load_issue_lock(
+        remote=remote, org=o, repo=r, issue_number=int(issue_number)
+    )
+    git_state = issue_lock_worktree.read_worktree_git_state(workspace)
+    registered = author_mutation_worktree.path_in_git_worktree_list(
+        workspace, PROJECT_ROOT
+    )
+
+    remote_probe = branch_publish.read_remote_branch_head(
+        workspace, git_remote, branch_name
+    )
+    ancestry = None
+    probe_head = (remote_probe.get("remote_head_sha") or "").strip()
+    if probe_head and probe_head.lower() != (expected_head or "").strip().lower():
+        ancestry = branch_publish.read_is_ancestor(workspace, probe_head, expected_head)
+
+    # A PR on this very branch is this work's own PR, not a rival claim. Only an
+    # open PR for the same issue on a *different* branch is a competing claim.
+    competing: list = []
+    for pull in _list_open_pulls(h, o, r, _auth(h)):
+        ref = str((pull.get("head") or {}).get("ref") or "")
+        if not ref or ref == branch_name:
+            continue
+        if issue_lock_adoption.branch_carries_issue_marker(ref, int(issue_number)):
+            competing.append(pull.get("number"))
+
+    observed_hashes = None
+    if expected_file_hashes:
+        observed_hashes = branch_publish.hash_worktree_files(
+            workspace, list(expected_file_hashes.keys())
+        )
+
+    claimant = _work_lease_claimant(h)
+    assessment = branch_publish.assess_unpublished_commit_publication(
+        existing_lock,
+        issue_number=int(issue_number),
+        branch_name=branch_name,
+        worktree_path=workspace,
+        expected_head=expected_head,
+        remote=remote,
+        org=o,
+        repo=r,
+        identity=claimant.get("username"),
+        profile=claimant.get("profile"),
+        worktree_state=git_state,
+        worktree_registered=registered,
+        remote_probe=remote_probe,
+        ancestry=ancestry,
+        competing_open_prs=competing,
+        expected_file_hashes=expected_file_hashes,
+        observed_file_hashes=observed_hashes,
+    )
+
+    if assessment["outcome"] == branch_publish.REFUSED:
+        return _publication_block(
+            assessment["reasons"], evidence=assessment["evidence"]
+        )
+
+    if dry_run:
+        return {
+            "success": True,
+            "performed": False,
+            "published": False,
+            "verified": False,
+            "dry_run": True,
+            "outcome": assessment["outcome"],
+            "would_publish": assessment["publish_sanctioned"],
+            "remote_head_sha": assessment["evidence"].get("remote_head_sha"),
+            "reasons": [],
+            "evidence": assessment["evidence"],
+            "issue_lock_record_mutated": False,
+            "workflow_lease_touched": False,
+        }
+
+    performed = False
+    if assessment["publish_sanctioned"]:
+        with _audited(
+            task, host=h, remote=remote, org=o, repo=r,
+            target_branch=branch_name,
+            request_metadata={
+                "issue_number": int(issue_number),
+                "expected_head": expected_head,
+                "git_remote": git_remote,
+            },
+        ):
+            push = branch_publish.publish_commit_to_remote_branch(
+                worktree_path=workspace,
+                remote_name=git_remote,
+                branch_name=branch_name,
+                expected_head=expected_head,
+            )
+        if not push.get("success"):
+            return _publication_block(
+                push.get("reasons") or ["publication failed"],
+                evidence=assessment["evidence"],
+                stderr=push.get("stderr"),
+            )
+        performed = True
+
+    verification = branch_publish.verify_published_head(
+        worktree_path=workspace,
+        remote_name=git_remote,
+        branch_name=branch_name,
+        expected_head=expected_head,
+    )
+    if not verification.get("verified"):
+        return _publication_block(
+            verification.get("reasons") or ["read-after-write verification failed"],
+            evidence=assessment["evidence"],
+            performed=performed,
+            published=performed,
+            remote_head_sha=verification.get("remote_head_sha"),
+        )
+
+    return {
+        "success": True,
+        "performed": performed,
+        "published": True,
+        "verified": True,
+        "outcome": assessment["outcome"],
+        "remote_head_sha": verification.get("remote_head_sha"),
+        "reasons": [],
+        "evidence": assessment["evidence"],
+        # Publication is the whole of this operation's authority (#812 AC23).
+        "issue_lock_record_mutated": False,
+        "workflow_lease_touched": False,
+        "exact_next_action": (
+            f"Remote head for '{branch_name}' is now observable. Call "
+            f"gitea_lock_issue(issue_number={int(issue_number)}, "
+            f"branch_name='{branch_name}', worktree_path='{workspace}') to renew "
+            "the exact-owner lease, then gitea_create_pr."
+        ),
+    }
+
+
 # Merge methods supported by the Gitea merge API.
 _MERGE_METHODS = ("merge", "squash", "rebase")
 
@@ -12536,10 +12784,39 @@ def _try_auto_switch_for_operation(op: str, host: str | None = None) -> bool:
     return False
 
 
+def _git_default_remote_name(root: str) -> str:
+    """First configured git remote name for *root*, defaulting to 'origin'.
+
+    Used to resolve the live remote master target for parity (#610). Best
+    effort: any failure falls back to 'origin' so callers never raise.
+    """
+    try:
+        res = subprocess.run(
+            ["git", "-C", root, "remote"],
+            capture_output=True, text=True, check=False,
+        )
+    except Exception:
+        return "origin"
+    if res.returncode != 0:
+        return "origin"
+    names = [n.strip() for n in (res.stdout or "").splitlines() if n.strip()]
+    return names[0] if names else "origin"
+
+
 def _current_master_parity() -> dict:
-    """Assess this process's code against the on-disk master HEAD (#420)."""
+    """Assess this process's code against local and live remote master (#420/#610).
+
+    Compares the daemon's startup commit, the on-disk checkout HEAD, and the
+    live remote master target. A stale daemon relative to live master fails
+    closed for mutations even when the local checkout HEAD still matches the
+    startup commit. The live-remote read is best effort: an unresolved live
+    head leaves read-only diagnostics unblocked but is never mutation-safe.
+    """
     current_head = master_parity_gate.read_git_head(PROJECT_ROOT)
-    return master_parity_gate.assess_master_parity(_STARTUP_PARITY, current_head)
+    live_head = master_parity_gate.read_remote_master_head(
+        PROJECT_ROOT, remote=_git_default_remote_name(PROJECT_ROOT))
+    return master_parity_gate.assess_master_parity(
+        _STARTUP_PARITY, current_head, live_remote_head=live_head)
 
 
 def _current_runtime_mode_report(refresh: bool = False) -> dict:
@@ -16480,15 +16757,34 @@ def gitea_get_runtime_context(
         "restart_required": parity["restart_required"],
         "startup_head": parity["startup_head"],
         "current_head": parity["current_head"],
+        # #610 distinguished mutation-safety signals:
+        "daemon_start_head": parity["daemon_start_head"],
+        "local_head": parity["local_head"],
+        "live_remote_head": parity["live_remote_head"],
+        "live_known": parity["live_known"],
+        "live_stale": parity["live_stale"],
+        "mutation_safe": parity["mutation_safe"],
         "summary": master_parity_gate.format_parity(parity),
         "mutation_gate_enforced": not master_parity_gate.gate_disabled(),
+        # #610: the capability resolver is authoritative for mutation safety;
+        # local parity alone must never authorize a mutation.
+        "resolver_authoritative_for_mutation_safety": True,
     }
-    if parity["stale"] and not master_parity_gate.gate_disabled():
-        safe_next_action = (
-            "Server code is stale relative to master; restart the Gitea MCP "
-            "server to load current capability gates before mutating. "
-            f"({master_parity_gate.format_parity(parity)})"
-        )
+    if parity["restart_required"] and not master_parity_gate.gate_disabled():
+        if parity["live_stale"]:
+            safe_next_action = (
+                "Daemon is stale relative to LIVE remote master "
+                f"(started {parity['startup_head'][:12] if parity['startup_head'] else 'unknown'}, "
+                f"live master {parity['live_remote_head'][:12] if parity['live_remote_head'] else 'unknown'}); "
+                "restart/reconnect the Gitea MCP server before mutating. The "
+                "capability resolver is authoritative for mutation safety."
+            )
+        else:
+            safe_next_action = (
+                "Server code is stale relative to master; restart the Gitea MCP "
+                "server to load current capability gates before mutating. "
+                f"({master_parity_gate.format_parity(parity)})"
+            )
         result["safe_next_action"] = safe_next_action
 
     if reveal and h:
@@ -16537,6 +16833,13 @@ def gitea_assess_master_parity(
         "determinable": parity["determinable"],
         "startup_head": parity["startup_head"],
         "current_head": parity["current_head"],
+        # #610 distinguished mutation-safety signals:
+        "daemon_start_head": parity["daemon_start_head"],
+        "local_head": parity["local_head"],
+        "live_remote_head": parity["live_remote_head"],
+        "live_known": parity["live_known"],
+        "live_stale": parity["live_stale"],
+        "mutation_safe": parity["mutation_safe"],
         "mutation_gate_enforced": enforced,
         "summary": master_parity_gate.format_parity(parity),
         "reasons": parity["reasons"],
@@ -16553,7 +16856,7 @@ def gitea_assess_master_parity(
             source=canonical_source,
         ),
     }
-    if parity["stale"] and enforced:
+    if parity["restart_required"] and enforced:
         out["report"] = master_parity_gate.parity_report(parity)
     return out
 

master_parity_gate.py

@@ -24,12 +24,50 @@ from __future__ import annotations
 
 import os
 import subprocess
+import time
+
+# Live-remote head cache: the parity gate runs on every mutation and every
+# runtime-context read, so the ``git ls-remote`` result is cached briefly to
+# avoid a network round-trip per call (#610). Keyed by (root, remote, branch).
+_REMOTE_HEAD_CACHE: dict[tuple[str, str, str], tuple[float, str | None]] = {}
+_REMOTE_HEAD_TTL = 60.0
+
+# When True, ``read_remote_master_head`` never performs ``git ls-remote`` unless
+# ``GITEA_TEST_LIVE_REMOTE_HEAD`` is set. Conftest enables this suite-wide so
+# feature worktrees (whose HEAD differs from live master) cannot flip legacy
+# runtime-context assertions to live_stale, and so unit tests never depend on
+# a live network (PR #788 F1/F2 / issue #610). Module-level (not env-only) so
+# ``patch.dict(os.environ, …, clear=True)`` cannot re-enable the probe.
+_HERMETIC_TEST_MODE: bool = False
+
+
+def _clear_remote_head_cache() -> None:
+    """Reset the live-remote head cache (test isolation / forced refresh)."""
+    _REMOTE_HEAD_CACHE.clear()
+
+
+def set_hermetic_test_mode(enabled: bool) -> None:
+    """Enable or disable suite-wide hermetic live-remote reads (tests only)."""
+    global _HERMETIC_TEST_MODE
+    _HERMETIC_TEST_MODE = bool(enabled)
+    _clear_remote_head_cache()
+
+
+def hermetic_test_mode() -> bool:
+    """Return whether hermetic live-remote reads are active."""
+    return bool(_HERMETIC_TEST_MODE)
+
 
 # Environment escape hatches (ops + tests):
 #   GITEA_MCP_DISABLE_PARITY_GATE  -> disable enforcement entirely (fail open).
 #   GITEA_TEST_CURRENT_HEAD        -> force the "current" HEAD read, for tests.
 ENV_DISABLE = "GITEA_MCP_DISABLE_PARITY_GATE"
 ENV_TEST_CURRENT_HEAD = "GITEA_TEST_CURRENT_HEAD"
+# GITEA_TEST_LIVE_REMOTE_HEAD -> force the live remote master read, for tests.
+ENV_TEST_LIVE_REMOTE_HEAD = "GITEA_TEST_LIVE_REMOTE_HEAD"
+# GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE -> opt a single test into a real ls-remote
+# even when hermetic mode is on (rare; prefer ENV_TEST_LIVE_REMOTE_HEAD).
+ENV_TEST_ALLOW_LIVE_REMOTE_PROBE = "GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE"
 
 
 def read_git_head(root: str) -> str | None:
@@ -58,6 +96,75 @@ def read_git_head(root: str) -> str | None:
     return (res.stdout or "").strip() or None
 
 
+def read_remote_master_head(
+    root: str,
+    remote: str = "origin",
+    branch: str = "master",
+    ttl: float = _REMOTE_HEAD_TTL,
+) -> str | None:
+    """Return the live remote ``branch`` commit SHA, or ``None`` (#610).
+
+    Resolves the *live* target commit via ``git ls-remote`` so parity can tell
+    a daemon that is behind the live remote master apart from one whose local
+    checkout simply hasn't been pulled. ``None`` means the live head could not
+    be resolved (offline, no such remote, git unavailable, error) -- callers
+    must treat unknown live state as *not mutation-safe* while never blocking
+    read-only diagnostics. A ``GITEA_TEST_LIVE_REMOTE_HEAD`` override takes
+    precedence so the wiring can be exercised deterministically and offline.
+
+    The result is cached for *ttl* seconds per (root, remote, branch) so the
+    gate does not run a network probe on every mutation/read (``ttl=0`` forces
+    a live probe). Both hits and ``None`` misses are cached to bound offline
+    latency; the env override bypasses the cache and the subprocess entirely.
+
+    Under suite hermetic mode (``set_hermetic_test_mode(True)``, set by
+    conftest) a missing override returns ``None`` without network I/O so
+    feature-worktree test runs cannot observe live_stale against real master
+    (PR #788 F1) and unit tests stay offline (F2). Opt out with an explicit
+    ``GITEA_TEST_LIVE_REMOTE_HEAD`` pin or ``GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE``.
+    """
+    forced = os.environ.get(ENV_TEST_LIVE_REMOTE_HEAD)
+    if forced is not None:
+        return forced.strip() or None
+    if _HERMETIC_TEST_MODE and not (
+        os.environ.get(ENV_TEST_ALLOW_LIVE_REMOTE_PROBE) or ""
+    ).strip():
+        # Hermetic default: live head unknown. live_stale stays False;
+        # mutation_safe is False when live is unknown (documented #610 note).
+        return None
+    # Defense in depth: even without the module flag, never probe while pytest
+    # is running unless the test opted into a real probe or set an override.
+    if (os.environ.get("PYTEST_CURRENT_TEST") or "").strip() and not (
+        os.environ.get(ENV_TEST_ALLOW_LIVE_REMOTE_PROBE) or ""
+    ).strip():
+        return None
+    if not root:
+        return None
+    key = (root, remote, branch)
+    now = time.monotonic()
+    if ttl > 0:
+        cached = _REMOTE_HEAD_CACHE.get(key)
+        if cached is not None and (now - cached[0]) < ttl:
+            return cached[1]
+    sha: str | None = None
+    try:
+        res = subprocess.run(
+            ["git", "-C", root, "ls-remote", remote, f"refs/heads/{branch}"],
+            capture_output=True,
+            text=True,
+            check=False,
+            timeout=5,
+        )
+        if res.returncode == 0:
+            lines = (res.stdout or "").strip().splitlines()
+            if lines:
+                sha = lines[0].split("\t", 1)[0].split()[0].strip() or None
+    except Exception:
+        sha = None
+    _REMOTE_HEAD_CACHE[key] = (now, sha)
+    return sha
+
+
 def capture_startup_parity(root: str, head: str | None = None) -> dict:
     """Capture the process source-tree baseline once at server startup.
 
@@ -72,18 +179,38 @@ def _short(sha: str | None) -> str:
     return sha[:12] if sha else "unknown"
 
 
-def assess_master_parity(startup: dict | None, current_head: str | None) -> dict:
+def assess_master_parity(
+    startup: dict | None,
+    current_head: str | None,
+    live_remote_head: str | None = None,
+) -> dict:
     """Compare the startup baseline against the current on-disk ``HEAD``.
 
-    Pure: both HEADs are supplied by the caller. Returns a structured result:
+    Pure: all HEADs are supplied by the caller. Returns a structured result:
 
     - ``in_parity``    -- server code matches the on-disk master (or parity
       could not be determined, which is not treated as stale).
     - ``stale``        -- the on-disk master has definitively advanced past the
       running process.
-    - ``restart_required`` -- alias of ``stale``; the recovery action.
-    - ``determinable`` -- whether both HEADs were known well enough to compare.
+    - ``restart_required`` -- ``stale`` or ``live_stale``; the recovery action.
+    - ``determinable`` -- whether both local HEADs were known well enough to
+      compare.
     - ``startup_head`` / ``current_head`` / ``reasons``.
+
+    #610 adds live-remote awareness so a daemon that is stale relative to the
+    *live* remote master cannot report a mutation-safe result even when the
+    local checkout HEAD still matches the daemon's startup commit:
+
+    - ``daemon_start_head`` -- the commit the running process started at
+      (alias of ``startup_head``, named for clarity in reports).
+    - ``local_head`` -- the on-disk checkout HEAD (alias of ``current_head``).
+    - ``live_remote_head`` -- the live remote target commit, or ``None`` when it
+      could not be fetched.
+    - ``live_known`` -- whether the live remote target was resolved.
+    - ``live_stale`` -- the live remote master has advanced past the running
+      process (daemon is behind live master) even if local parity is green.
+    - ``mutation_safe`` -- the daemon code, local checkout, and live remote
+      target all agree; the only state in which a mutation may rely on parity.
     """
     startup_head = (startup or {}).get("startup_head")
     reasons: list[str] = []
@@ -91,32 +218,56 @@ def assess_master_parity(startup: dict | None, current_head: str | None) -> dict
     if startup_head is None:
         reasons.append(
             "startup commit was not captured; code parity cannot be enforced")
-        return _result(True, False, False, startup_head, current_head, reasons)
+        return _result(True, False, False, startup_head, current_head,
+                       live_remote_head, False, reasons)
 
     if current_head is None:
         reasons.append(
             "current workspace HEAD could not be read; code parity cannot be "
             "enforced")
-        return _result(True, False, False, startup_head, current_head, reasons)
+        return _result(True, False, False, startup_head, current_head,
+                       live_remote_head, False, reasons)
 
-    if startup_head == current_head:
-        return _result(True, False, True, startup_head, current_head, reasons)
+    local_in_parity = startup_head == current_head
+    local_stale = not local_in_parity
+    if local_stale:
+        reasons.append(
+            f"MCP server started at commit {_short(startup_head)} but the "
+            f"workspace master is now {_short(current_head)}; restart the "
+            f"server to load the current capability gates")
 
-    reasons.append(
-        f"MCP server started at commit {_short(startup_head)} but the workspace "
-        f"master is now {_short(current_head)}; restart the server to load the "
-        f"current capability gates")
-    return _result(False, True, True, startup_head, current_head, reasons)
+    live_known = live_remote_head is not None
+    live_stale = live_known and live_remote_head != startup_head
+    if live_stale:
+        reasons.append(
+            f"live remote master is {_short(live_remote_head)} but the MCP "
+            f"server started at {_short(startup_head)}; the daemon is stale "
+            f"relative to live master -- restart/reconnect before mutating")
+
+    return _result(
+        local_in_parity, local_stale, True, startup_head, current_head,
+        live_remote_head, live_stale, reasons)
 
 
-def _result(in_parity, stale, determinable, startup_head, current_head, reasons):
+def _result(in_parity, stale, determinable, startup_head, current_head,
+            live_remote_head, live_stale, reasons):
+    live_known = live_remote_head is not None
+    mutation_safe = (
+        determinable and in_parity and live_known and not live_stale)
     return {
         "in_parity": in_parity,
         "stale": stale,
-        "restart_required": stale,
+        "restart_required": stale or live_stale,
         "determinable": determinable,
         "startup_head": startup_head,
         "current_head": current_head,
+        # #610 distinguished signals:
+        "daemon_start_head": startup_head,
+        "local_head": current_head,
+        "live_remote_head": live_remote_head,
+        "live_known": live_known,
+        "live_stale": live_stale,
+        "mutation_safe": mutation_safe,
         "reasons": list(reasons),
     }
 
@@ -130,11 +281,13 @@ def parity_block_reasons(assessment: dict) -> list[str]:
     """Block reasons for a mutation gate (empty when the mutation may proceed).
 
     A disabled gate or an in-parity / non-determinable assessment yields no
-    reasons; only a definitively stale server blocks.
+    reasons. A definitively stale server blocks, and (#610) a daemon that is
+    stale relative to the *live* remote master blocks even when the local
+    checkout HEAD still matches the daemon's startup commit.
     """
     if gate_disabled():
         return []
-    if assessment.get("stale"):
+    if assessment.get("stale") or assessment.get("live_stale"):
         return list(assessment.get("reasons") or
                     ["server code is stale relative to master (fail closed)"])
     return []
@@ -147,6 +300,10 @@ def parity_report(assessment: dict) -> dict:
         "restart_required": True,
         "startup_head": assessment.get("startup_head"),
         "current_head": assessment.get("current_head"),
+        # #610: name the live remote target so the report distinguishes a
+        # local-code stale from a daemon-behind-live-master stale.
+        "live_remote_head": assessment.get("live_remote_head"),
+        "live_stale": bool(assessment.get("live_stale")),
         "reasons": list(assessment.get("reasons") or []),
         "recovery": [
             "The running MCP server is executing code older than the current "
@@ -157,6 +314,45 @@ def parity_report(assessment: dict) -> dict:
     }
 
 
+def parity_resolver_disagreement(
+    assessment: dict,
+    resolver_restart_required: bool,
+) -> dict | None:
+    """Typed blocker when the resolver requires restart but parity looks green.
+
+    The capability resolver (``gitea_resolve_task_capability``) detects stale
+    runtime authoritatively for mutation safety (#610). When it requires a
+    restart, local-only parity must never override it: this returns a typed,
+    fail-closed blocker that names the resolver as authoritative. Returns
+    ``None`` when the resolver does not require a restart.
+    """
+    if not resolver_restart_required:
+        return None
+    parity_optimistic = bool(assessment.get("in_parity")) and not (
+        assessment.get("stale") or assessment.get("live_stale"))
+    return {
+        "kind": "parity_resolver_disagreement",
+        "restart_required": True,
+        "resolver_authoritative": True,
+        "parity_optimistic": parity_optimistic,
+        "daemon_start_head": assessment.get("daemon_start_head"),
+        "local_head": assessment.get("local_head"),
+        "live_remote_head": assessment.get("live_remote_head"),
+        "reasons": [
+            "The capability resolver requires a restart/reconnect (stale "
+            "runtime) but master-parity reported local code as in-parity. "
+            "The resolver is authoritative for mutation safety; do not mutate "
+            "on local parity alone. Restart/reconnect the Gitea MCP server "
+            "and re-verify before mutating.",
+        ],
+        "recovery": [
+            "Trust the resolver: treat this session as stale.",
+            "Restart or /mcp reconnect the Gitea MCP namespace so it reloads "
+            "current master and live target state, then re-run preflight.",
+        ],
+    }
+
+
 def format_parity(assessment: dict) -> str:
     """One-line human summary for logs / runtime context."""
     if assessment.get("stale"):

task_capability_map.py

@@ -62,6 +62,14 @@ TASK_CAPABILITY_MAP: dict[str, dict[str, str]] = {
         "permission": "gitea.branch.push",
         "role": "author",
     },
+    # #812 AC20: publish an already-committed, unpublished local head so
+    # exact-owner lease renewal has an observable remote head to reason about.
+    # Same authority as any other author push — deliberately not a new
+    # operation name, so it cannot widen an already-configured author profile.
+    "publish_unpublished_branch": {
+        "permission": "gitea.branch.push",
+        "role": "author",
+    },
     "create_pr": {
         "permission": "gitea.pr.create",
         "role": "author",
@@ -500,6 +508,7 @@ ROLE_EXCLUSIVE_TASKS: frozenset[str] = frozenset(
         "gitea_release_merger_pr_lease",
         "create_branch",
         "push_branch",
+        "publish_unpublished_branch",
         "create_pr",
         "commit_files",
         "gitea_commit_files",

tests/conftest.py

@@ -167,6 +167,35 @@ def _reset_mutation_authority(monkeypatch):
 import pytest
 
 
+@pytest.fixture(autouse=True)
+def _hermetic_live_remote_master_head():
+    """#610 / PR #788 F1/F2: keep live-remote parity reads offline in tests.
+
+    ``read_remote_master_head`` would otherwise ``git ls-remote`` whenever
+    ``GITEA_TEST_LIVE_REMOTE_HEAD`` is unset. Feature worktrees under
+    ``branches/`` always differ from live master, so legacy suites that assert
+    runtime-context ``safe_next_action`` flip to live_stale. Module-level
+    hermetic mode survives ``patch.dict(os.environ, …, clear=True)``.
+    Tests that exercise the real probe path call
+    ``master_parity_gate.set_hermetic_test_mode(False)`` and/or set
+    ``GITEA_TEST_ALLOW_LIVE_REMOTE_PROBE``.
+    """
+    try:
+        import master_parity_gate as _mpg
+
+        _mpg.set_hermetic_test_mode(True)
+    except Exception:
+        _mpg = None
+    try:
+        yield
+    finally:
+        if _mpg is not None:
+            try:
+                _mpg.set_hermetic_test_mode(False)
+            except Exception:
+                pass
+
+
 @pytest.fixture(autouse=True)
 def _deterministic_workspace_remotes():
     try:

tests/test_issue_812_publish_unpublished_commit.py

@@ -0,0 +1,650 @@
+"""Publication of an unpublished local commit (#812 AC20).
+
+Entry point B of #812: a registered worktree, clean, on its issue branch,
+holding a local commit that has never been published. Exact-owner lease renewal
+refuses such a claim for want of an observable remote head, and every existing
+publication path is lock-derived, so the two predicates close a cycle around
+work that is otherwise complete.
+
+These tests exercise the disposition through its *evidence*, never through any
+particular issue number: every case uses an arbitrary issue number against a
+synthetic repository, and the same assertions hold for any other. Nothing here
+reads, writes, or references the live protected worktree named in #812 AC17 —
+that content is preserved evidence for the duration of this work, so the
+fixtures below build their own repositories from scratch.
+
+The remote is a local bare repository, so publication and read-after-write
+verification are genuinely executed rather than mocked.
+"""
+
+from __future__ import annotations
+
+import os
+import subprocess
+import sys
+import tempfile
+import unittest
+from datetime import datetime, timedelta, timezone
+from unittest.mock import patch
+
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+
+import branch_publish  # noqa: E402
+import issue_lock_provenance  # noqa: E402
+import issue_lock_renewal  # noqa: E402
+import issue_lock_store  # noqa: E402
+import mcp_server  # noqa: E402
+from mutation_profile_fixture import shared_mutation_env  # noqa: E402
+
+ISSUE = 9812
+BRANCH = f"feat/issue-{ISSUE}-publish-fixture"
+IDENTITY = "example-user"
+PROFILE = "test-author-prgs"
+ORG = "Scaled-Tech-Consulting"
+REPO = "Gitea-Tools"
+GIT_REMOTE = "prgs"
+
+
+def _ts(hours: int) -> str:
+    return (
+        (datetime.now(timezone.utc) + timedelta(hours=hours))
+        .isoformat()
+        .replace("+00:00", "Z")
+    )
+
+
+class _PublishBase(unittest.TestCase):
+    """Real git repo + real bare remote + durable lock naming the caller.
+
+    The recorded owner pid is deliberately **this live process**. That mirrors
+    the production shape #812 documents, where the pid belongs to a long-running
+    MCP daemon rather than to a dead author client, and it proves publication
+    never depends on a dead process (#812 AC24).
+    """
+
+    def setUp(self):
+        self.lock_dir = tempfile.TemporaryDirectory()
+        self.addCleanup(self.lock_dir.cleanup)
+        self.origin = tempfile.mkdtemp(prefix="issue812-origin-")
+        self.repo = tempfile.mkdtemp(prefix="issue812-work-")
+        for path in (self.origin, self.repo):
+            self.addCleanup(
+                lambda p=path: subprocess.run(["rm", "-rf", p], check=False)
+            )
+        self._init_repos()
+        self.remotes = patch.dict(
+            mcp_server.REMOTES,
+            {"prgs": {"host": "gitea.prgs.cc", "org": ORG, "repo": REPO}},
+        )
+        self.remotes.start()
+        self.addCleanup(patch.stopall)
+        mcp_server._IDENTITY_CACHE.clear()
+
+    # ── fixture construction ─────────────────────────────────────────────
+    def _git(self, *args, cwd=None):
+        return subprocess.run(
+            ["git", "-C", cwd or self.repo, *args],
+            capture_output=True,
+            text=True,
+            check=True,
+        )
+
+    def _init_repos(self):
+        subprocess.run(
+            ["git", "init", "-q", "--bare", "-b", "master", self.origin], check=True
+        )
+        self._git("init", "-q", "-b", "master")
+        self._git("config", "user.email", "[email protected]")
+        self._git("config", "user.name", "Test")
+        self._git("remote", "add", GIT_REMOTE, self.origin)
+
+        with open(os.path.join(self.repo, "seed.txt"), "w") as fh:
+            fh.write("seed\n")
+        self._git("add", "seed.txt")
+        self._git("commit", "-q", "-m", "seed")
+        self.base_sha = self._git("rev-parse", "HEAD").stdout.strip()
+        self._git("push", "-q", GIT_REMOTE, "master")
+
+        self._git("checkout", "-q", "-b", BRANCH)
+        with open(os.path.join(self.repo, "work.txt"), "w") as fh:
+            fh.write("unpublished implementation\n")
+        self._git("add", "work.txt")
+        self._git("commit", "-q", "-m", "unpublished implementation")
+        self.head_sha = self._git("rev-parse", "HEAD").stdout.strip()
+        self.worktree = os.path.realpath(self.repo)
+
+    def lock_path(self):
+        return issue_lock_store.lock_file_path(
+            remote="prgs", org=ORG, repo=REPO, issue_number=ISSUE,
+            lock_dir=self.lock_dir.name,
+        )
+
+    def write_lock(self, **overrides):
+        path = self.lock_path()
+        claimant = overrides.pop(
+            "claimant", {"username": IDENTITY, "profile": PROFILE}
+        )
+        pid = overrides.pop("session_pid", os.getpid())
+        lease = {
+            "operation_type": issue_lock_store.AUTHOR_ISSUE_WORK_LEASE,
+            "issue_number": ISSUE,
+            "pr_number": None,
+            "branch": overrides.get("branch_name", BRANCH),
+            "worktree_path": overrides.get("worktree_path", self.worktree),
+            "claimant": claimant,
+            "created_at": _ts(-2),
+            "last_heartbeat_at": _ts(-2),
+            # Expired: entry point B's lease has lapsed, which is precisely why
+            # renewal — and therefore a published head — is needed.
+            "expires_at": _ts(-1),
+        }
+        lease.update(overrides.pop("work_lease", {}))
+        data = {
+            "issue_number": ISSUE,
+            "branch_name": BRANCH,
+            "remote": "prgs",
+            "org": ORG,
+            "repo": REPO,
+            "worktree_path": self.worktree,
+            "session_pid": pid,
+            "pid": pid,
+            "lock_generation": 1,
+            "work_lease": lease,
+            "lock_provenance": issue_lock_provenance.build_sanctioned_lock_provenance(
+                tool="gitea_lock_issue", claimant=claimant
+            ),
+        }
+        data.update(overrides)
+        data["lock_file_path"] = path
+        issue_lock_store.save_lock_file(path, data)
+        return path
+
+    def _tool_env(self):
+        env = shared_mutation_env(
+            PROFILE, include_example_repo=True,
+            GITEA_ISSUE_LOCK_DIR=self.lock_dir.name,
+        )
+        env["GITEA_ISSUE_LOCK_DIR"] = self.lock_dir.name
+        # These tests repoint PROJECT_ROOT at a synthetic repository so the
+        # registered-worktree proof runs for real. Pin the parity gate to the
+        # server's own startup head so that repointing does not read as a stale
+        # daemon; the gate itself stays live and enforced.
+        startup_head = mcp_server._STARTUP_PARITY.get("startup_head") or ""
+        env["GITEA_TEST_CURRENT_HEAD"] = startup_head
+        env["GITEA_TEST_LIVE_REMOTE_HEAD"] = startup_head
+        return env
+
+    # ── tool driver ──────────────────────────────────────────────────────
+    def run_publish(self, *, open_prs=None, expected_head=None, **kwargs):
+        """Drive the public publication tool against the synthetic fixture."""
+        env = self._tool_env()
+        with patch(
+            "mcp_server._list_open_pulls", return_value=list(open_prs or [])
+        ), patch(
+            "mcp_server._auth", return_value="token x"
+        ), patch(
+            "mcp_server.get_auth_header", return_value="token x"
+        ), patch(
+            "mcp_server._work_lease_claimant",
+            return_value={"username": IDENTITY, "profile": PROFILE},
+        ), patch.object(
+            mcp_server, "PROJECT_ROOT", self.repo
+        ), patch.dict(os.environ, env, clear=True):
+            os.environ["GITEA_ISSUE_LOCK_DIR"] = self.lock_dir.name
+            return mcp_server.gitea_publish_unpublished_issue_branch(
+                issue_number=kwargs.pop("issue_number", ISSUE),
+                branch_name=kwargs.pop("branch_name", BRANCH),
+                worktree_path=kwargs.pop("worktree_path", self.worktree),
+                expected_head=expected_head or self.head_sha,
+                remote="prgs",
+                git_remote_name=kwargs.pop("git_remote_name", GIT_REMOTE),
+                **kwargs,
+            )
+
+    def remote_head(self, branch=BRANCH):
+        res = subprocess.run(
+            ["git", "-C", self.origin, "rev-parse", "--verify", "--quiet", branch],
+            capture_output=True, text=True, check=False,
+        )
+        return (res.stdout or "").strip() or None
+
+
+class TestSuccessfulPublication(_PublishBase):
+    """AC20 — the branch becomes observable and is verified after the write."""
+
+    def test_publishes_clean_unpublished_commit(self):
+        self.write_lock()
+        self.assertIsNone(self.remote_head(), "fixture must start unpublished")
+
+        result = self.run_publish()
+
+        self.assertTrue(result["success"], result.get("reasons"))
+        self.assertTrue(result["performed"])
+        self.assertTrue(result["published"])
+        self.assertTrue(result["verified"], "read-after-write must be proven")
+        self.assertEqual(result["remote_head_sha"], self.head_sha)
+        self.assertEqual(self.remote_head(), self.head_sha)
+
+    def test_publication_does_not_rewrite_the_commit(self):
+        self.write_lock()
+        self.run_publish()
+        # The published object is the same commit, not a copy or a rewrite.
+        self.assertEqual(self.remote_head(), self.head_sha)
+        self.assertEqual(
+            self._git("rev-parse", "HEAD").stdout.strip(), self.head_sha
+        )
+
+    def test_exact_next_action_names_the_lock_call(self):
+        self.write_lock()
+        result = self.run_publish()
+        self.assertIn("gitea_lock_issue", result["exact_next_action"])
+
+
+class TestFailsClosed(_PublishBase):
+    """AC20/AC9 — each refusal reason, exercised independently."""
+
+    def test_changed_local_head_refuses(self):
+        self.write_lock()
+        stale = self.base_sha  # a real commit, but not the declared head
+        result = self.run_publish(expected_head=stale)
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("local commit changed" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+        self.assertIsNone(self.remote_head(), "refusal must not publish")
+
+    def test_abbreviated_sha_refuses(self):
+        self.write_lock()
+        result = self.run_publish(expected_head=self.head_sha[:8])
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("40-character" in r for r in result["reasons"]), result["reasons"]
+        )
+
+    def test_dirty_tracked_worktree_refuses(self):
+        self.write_lock()
+        with open(os.path.join(self.repo, "work.txt"), "a") as fh:
+            fh.write("uncommitted edit\n")
+
+        result = self.run_publish()
+
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("dirty tracked files" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+        self.assertIn("work.txt", result["evidence"]["dirty_tracked_files"])
+        self.assertIsNone(self.remote_head())
+
+    def test_untracked_file_refuses(self):
+        self.write_lock()
+        with open(os.path.join(self.repo, "stray.txt"), "w") as fh:
+            fh.write("not committed\n")
+
+        result = self.run_publish()
+
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("untracked files" in r for r in result["reasons"]), result["reasons"]
+        )
+        self.assertIn("stray.txt", result["evidence"]["untracked_files"])
+        self.assertIsNone(self.remote_head())
+
+    def test_unexpected_remote_head_refuses(self):
+        """A remote head that is not an ancestor must never be overwritten."""
+        self.write_lock()
+        # Publish a divergent commit to the branch from a separate line.
+        self._git("checkout", "-q", "-b", "divergent", self.base_sha)
+        with open(os.path.join(self.repo, "other.txt"), "w") as fh:
+            fh.write("someone else's work\n")
+        self._git("add", "other.txt")
+        self._git("commit", "-q", "-m", "divergent")
+        divergent = self._git("rev-parse", "HEAD").stdout.strip()
+        self._git("push", "-q", GIT_REMOTE, f"{divergent}:refs/heads/{BRANCH}")
+        self._git("checkout", "-q", BRANCH)
+
+        result = self.run_publish()
+
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("not an ancestor" in r for r in result["reasons"]), result["reasons"]
+        )
+        self.assertEqual(
+            self.remote_head(), divergent, "the other head must survive intact"
+        )
+
+    def test_fast_forward_remote_head_is_allowed(self):
+        """An ancestor head is an honest fast-forward, not a conflict."""
+        self.write_lock()
+        self._git("push", "-q", GIT_REMOTE, f"{self.base_sha}:refs/heads/{BRANCH}")
+
+        result = self.run_publish()
+
+        self.assertTrue(result["success"], result.get("reasons"))
+        self.assertTrue(result["evidence"]["fast_forward_from_remote"])
+        self.assertEqual(self.remote_head(), self.head_sha)
+
+    def test_content_hash_mismatch_refuses(self):
+        self.write_lock()
+        wrong = {"work.txt": "0" * 64}
+
+        result = self.run_publish(expected_file_hashes=wrong)
+
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("declared content hashes" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+        self.assertFalse(result["evidence"]["file_hashes_verified"])
+        self.assertIsNone(self.remote_head())
+
+    def test_matching_content_hashes_publish(self):
+        self.write_lock()
+        digests = branch_publish.hash_worktree_files(self.worktree, ["work.txt"])
+
+        result = self.run_publish(expected_file_hashes=digests)
+
+        self.assertTrue(result["success"], result.get("reasons"))
+        self.assertTrue(result["evidence"]["file_hashes_verified"])
+
+    def test_missing_declared_file_refuses(self):
+        self.write_lock()
+        result = self.run_publish(expected_file_hashes={"absent.txt": "0" * 64})
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("missing or unreadable" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+
+    def test_foreign_claimant_refuses(self):
+        """Ownership comes from the durable record, not from the caller."""
+        self.write_lock(claimant={"username": "someone-else", "profile": PROFILE})
+
+        result = self.run_publish()
+
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("foreign claim" in r for r in result["reasons"]), result["reasons"]
+        )
+        self.assertIsNone(self.remote_head())
+
+    def test_foreign_profile_refuses(self):
+        self.write_lock(
+            claimant={"username": IDENTITY, "profile": "test-reviewer-prgs"}
+        )
+        result = self.run_publish()
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("claimant profile" in r for r in result["reasons"]), result["reasons"]
+        )
+
+    def test_absent_lock_record_refuses(self):
+        """No recorded claim means this cannot be used to bypass the lock."""
+        result = self.run_publish()  # no write_lock()
+
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("no durable issue-lock record" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+        self.assertIsNone(self.remote_head())
+
+    def test_branch_mismatch_against_lock_refuses(self):
+        self.write_lock(branch_name=f"feat/issue-{ISSUE}-different")
+        result = self.run_publish()
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("records branch" in r for r in result["reasons"]), result["reasons"]
+        )
+
+    def test_worktree_mismatch_against_lock_refuses(self):
+        self.write_lock(worktree_path="/tmp/some/other/worktree")
+        result = self.run_publish()
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("records worktree" in r for r in result["reasons"]), result["reasons"]
+        )
+
+    def test_competing_open_pr_on_another_branch_refuses(self):
+        self.write_lock()
+        competing = [{"number": 4242, "head": {"ref": f"fix/issue-{ISSUE}-rival"}}]
+
+        result = self.run_publish(open_prs=competing)
+
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("already claim issue" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+        self.assertIsNone(self.remote_head())
+
+    def test_open_pr_on_the_same_branch_is_not_competing(self):
+        """This branch's own PR is not a rival claim against itself."""
+        self.write_lock()
+        own = [{"number": 77, "head": {"ref": BRANCH}}]
+
+        result = self.run_publish(open_prs=own)
+
+        self.assertTrue(result["success"], result.get("reasons"))
+
+
+class TestGuardStrictnessPreserved(_PublishBase):
+    """AC15 — publication is an operation, never a weakening of the guards."""
+
+    def test_non_issue_branch_refuses(self):
+        self._git("checkout", "-q", "-b", "scratch/not-issue-linked")
+        self.write_lock(branch_name="scratch/not-issue-linked")
+
+        result = self.run_publish(branch_name="scratch/not-issue-linked")
+
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("issue-linked" in r for r in result["reasons"]), result["reasons"]
+        )
+
+    def test_stable_branch_refuses(self):
+        self.write_lock(branch_name="master")
+        result = self.run_publish(branch_name="master")
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("issue-linked" in r or "stable branch" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+
+    def test_branch_number_must_match_the_issue(self):
+        other = "feat/issue-7777-mismatched"
+        self._git("checkout", "-q", "-b", other)
+        self.write_lock(branch_name=other)
+        result = self.run_publish(branch_name=other)
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("does not carry issue number" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+
+    def test_unregistered_worktree_refuses(self):
+        """#713 — an improvised directory is not a registered worktree."""
+        path = self.write_lock()
+        assessment = branch_publish.assess_unpublished_commit_publication(
+            issue_lock_store.read_lock_file(path),
+            issue_number=ISSUE, branch_name=BRANCH, worktree_path=self.worktree,
+            expected_head=self.head_sha, remote="prgs", org=ORG, repo=REPO,
+            identity=IDENTITY, profile=PROFILE,
+            worktree_state={
+                "current_branch": BRANCH, "porcelain_status": "",
+                "head_sha": self.head_sha,
+            },
+            worktree_registered=False,
+            remote_probe={"probe_ok": True, "remote_branch_exists": False},
+        )
+        self.assertEqual(assessment["outcome"], branch_publish.REFUSED)
+        self.assertTrue(
+            any("not listed in git worktree list" in r
+                for r in assessment["reasons"]),
+            assessment["reasons"],
+        )
+
+    def test_unobservable_remote_refuses(self):
+        """An unknown remote state must not be mistaken for an absent branch."""
+        self.write_lock()
+        result = self.run_publish(git_remote_name="no-such-remote")
+        self.assertFalse(result["success"])
+        self.assertTrue(
+            any("could not be observed" in r for r in result["reasons"]),
+            result["reasons"],
+        )
+
+
+class TestRecordSeparation(_PublishBase):
+    """AC23 — the durable issue lock and the workflow lease are distinct."""
+
+    def test_publication_leaves_the_issue_lock_byte_identical(self):
+        path = self.write_lock()
+        with open(path, "rb") as fh:
+            before = fh.read()
+
+        result = self.run_publish()
+
+        self.assertTrue(result["success"], result.get("reasons"))
+        with open(path, "rb") as fh:
+            after = fh.read()
+        self.assertEqual(before, after, "publication must not mutate the lock record")
+        self.assertFalse(result["issue_lock_record_mutated"])
+        self.assertFalse(result["workflow_lease_touched"])
+
+    def test_refusal_also_reports_untouched_records(self):
+        result = self.run_publish()  # refuses: no lock record
+        self.assertFalse(result["issue_lock_record_mutated"])
+        self.assertFalse(result["workflow_lease_touched"])
+
+    def test_lock_generation_is_not_advanced(self):
+        path = self.write_lock()
+        self.run_publish()
+        lock = issue_lock_store.read_lock_file(path)
+        self.assertEqual(lock["lock_generation"], 1)
+
+
+class TestTruthfulProcessEvidence(_PublishBase):
+    """AC24 — a live daemon pid is never represented as a dead process."""
+
+    def test_live_recorded_pid_does_not_block_publication(self):
+        # The recorded pid is this live process, standing in for the live MCP
+        # daemon. Reclaim would refuse here; publication legitimately does not.
+        path = self.write_lock(session_pid=os.getpid())
+        lock = issue_lock_store.read_lock_file(path)
+        self.assertEqual(lock["pid"], os.getpid())
+
+        result = self.run_publish()
+
+        self.assertTrue(result["success"], result.get("reasons"))
+        self.assertEqual(self.remote_head(), self.head_sha)
+
+    def test_liveness_is_not_consulted_as_evidence(self):
+        self.write_lock(session_pid=os.getpid())
+        result = self.run_publish()
+        self.assertFalse(result["evidence"]["owner_pid_liveness_consulted"])
+
+    def test_reclaim_still_refuses_for_the_same_live_pid(self):
+        """Publication does not soften the reclaim predicate it routes around."""
+        path = self.write_lock(session_pid=os.getpid())
+        lock = issue_lock_store.read_lock_file(path)
+        reclaim = issue_lock_store.assess_expired_lock_reclaim(lock)
+        self.assertFalse(reclaim["reclaim_allowed"])
+
+
+class TestIdempotentRetry(_PublishBase):
+    """AC20 — retry is safe and read-after-write is proven every time."""
+
+    def test_second_publication_reports_already_published(self):
+        self.write_lock()
+        first = self.run_publish()
+        self.assertTrue(first["performed"])
+
+        second = self.run_publish()
+
+        self.assertTrue(second["success"], second.get("reasons"))
+        self.assertFalse(second["performed"], "no second push is needed")
+        self.assertTrue(second["published"])
+        self.assertTrue(second["verified"])
+        self.assertEqual(second["outcome"], branch_publish.ALREADY_PUBLISHED)
+        self.assertEqual(self.remote_head(), self.head_sha)
+
+
+class TestDryRun(_PublishBase):
+    """AC12 — dry run reports the decision and mutates nothing."""
+
+    def test_dry_run_reports_intent_without_publishing(self):
+        self.write_lock()
+
+        result = self.run_publish(dry_run=True)
+
+        self.assertTrue(result["success"])
+        self.assertTrue(result["dry_run"])
+        self.assertTrue(result["would_publish"])
+        self.assertFalse(result["performed"])
+        self.assertIsNone(self.remote_head(), "dry run must not publish")
+
+    def test_dry_run_and_apply_agree_on_a_refusal(self):
+        """AC11 — the reported decision does not depend on which mode ran."""
+        self.write_lock(claimant={"username": "someone-else", "profile": PROFILE})
+
+        dry = self.run_publish(dry_run=True)
+        applied = self.run_publish()
+
+        self.assertFalse(dry["success"])
+        self.assertFalse(applied["success"])
+        self.assertEqual(dry["reasons"], applied["reasons"])
+
+
+class TestRenewalUnblocked(_PublishBase):
+    """AC20/AC21 — renewal is permitted only after verified publication."""
+
+    def _renewal(self, remote_head):
+        return issue_lock_renewal.assess_exact_owner_lease_renewal(
+            issue_lock_store.read_lock_file(self.lock_path()),
+            issue_number=ISSUE, branch_name=BRANCH, worktree_path=self.worktree,
+            remote="prgs", org=ORG, repo=REPO,
+            identity=IDENTITY, profile=PROFILE,
+            current_branch=BRANCH, porcelain_status="", worktree_exists=True,
+            head_sha=self.head_sha, remote_head_sha=remote_head,
+        )
+
+    def test_renewal_refuses_before_publication(self):
+        self.write_lock()
+        decision = self._renewal(None)
+        self.assertFalse(decision["renewal_sanctioned"])
+        self.assertTrue(
+            any("unpublished branch" in r for r in decision["reasons"]),
+            decision["reasons"],
+        )
+
+    def test_renewal_is_sanctioned_after_publication(self):
+        self.write_lock()
+        result = self.run_publish()
+        self.assertTrue(result["verified"], result.get("reasons"))
+
+        decision = self._renewal(self.remote_head())
+
+        self.assertTrue(decision["renewal_sanctioned"], decision["reasons"])
+
+
+class TestProtectedAssetUntouched(unittest.TestCase):
+    """AC17 — no test or fixture may reference the protected worktree."""
+
+    def test_no_reference_to_the_protected_worktree(self):
+        here = os.path.dirname(os.path.abspath(__file__))
+        root = os.path.dirname(here)
+        needle = "issue-635-project-registry" + "-api"
+        for path in (
+            os.path.join(here, "test_issue_812_publish_unpublished_commit.py"),
+            os.path.join(root, "branch_publish.py"),
+        ):
+            with open(path, "r", encoding="utf-8") as fh:
+                body = fh.read()
+            self.assertNotIn(needle, body)
+
+
+if __name__ == "__main__":  # pragma: no cover
+    unittest.main()

tests/test_issue_815_preflight_worktree_forwarding.py

@@ -0,0 +1,622 @@
+"""Publication preflight must receive the caller's worktree (#815).
+
+``gitea_publish_unpublished_issue_branch`` takes a **required** ``worktree_path``
+but resolved it only *after* ``verify_preflight_purity`` had already run. Every
+workspace-resolution layer behind that preflight — canonical root, root checkout,
+create-issue bootstrap, the #618 branches-only guard, issue scope, and anti-stomp
+— therefore received ``None`` and fell back to the MCP process root. A daemon
+rooted at the stable control checkout refused a valid registered issue worktree
+that the caller had explicitly supplied, before the publication assessor ever ran.
+
+The #812 suite could not see this. Its fixture sets ``self.worktree =
+os.path.realpath(self.repo)`` and patches ``PROJECT_ROOT`` to that same path, so
+the fallback resolved to the very worktree the argument named. The production
+topology — control checkout on a stable branch, issue worktree somewhere else —
+was never constructed, and preflight additionally no-ops under pytest unless
+production guards are forced on.
+
+These tests build that topology honestly:
+
+* ``PROJECT_ROOT`` is a control checkout sitting on ``master``;
+* the registered issue worktree is a genuinely separate path under ``branches/``;
+* ``GITEA_TEST_FORCE_PRODUCTION_GUARDS`` is set so the #618 guard really runs;
+* no patch makes the issue worktree appear to be ``PROJECT_ROOT``.
+
+Nothing here reads, writes, or references the protected worktree named in #812
+AC17 and #815 AC9. Every fixture is built from scratch against a local bare
+remote, so publication and read-after-write verification genuinely execute.
+"""
+
+from __future__ import annotations
+
+import os
+import subprocess
+import sys
+import tempfile
+import unittest
+from datetime import datetime, timedelta, timezone
+from unittest.mock import patch
+
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+
+import issue_lock_provenance  # noqa: E402
+import issue_lock_store  # noqa: E402
+import mcp_server  # noqa: E402
+from mutation_profile_fixture import shared_mutation_env  # noqa: E402
+
+ISSUE = 9815
+BRANCH = f"feat/issue-{ISSUE}-forwarding-fixture"
+WORKTREE_DIRNAME = BRANCH.replace("/", "-")
+IDENTITY = "example-user"
+PROFILE = "test-author-prgs"
+ORG = "Scaled-Tech-Consulting"
+REPO = "Gitea-Tools"
+GIT_REMOTE = "prgs"
+
+AUTHOR_PROFILE = {
+    "profile_name": "prgs-author",
+    "role": "author",
+    "allowed_operations": [
+        "gitea.read", "gitea.issue.create", "gitea.issue.comment",
+        "gitea.pr.create", "gitea.repo.commit", "gitea.branch.push",
+    ],
+    "forbidden_operations": [],
+    "audit_label": "prgs-author",
+}
+
+
+def _ts(hours: int) -> str:
+    return (
+        (datetime.now(timezone.utc) + timedelta(hours=hours))
+        .isoformat()
+        .replace("+00:00", "Z")
+    )
+
+
+class TestPreflightReceivesTheWorktree(unittest.TestCase):
+    """AC1 — the supplied path reaches ``verify_preflight_purity`` itself.
+
+    Follows the #735 capture pattern: replace preflight with a recorder that
+    raises, so the argument can be proven forwarded without performing the
+    mutation. This is the direct unit-level statement of the defect.
+    """
+
+    def _capture_preflight(self, **kwargs):
+        captured: dict = {}
+
+        def _capture(*a, **kw):
+            captured.update(kw)
+            captured["_args"] = a
+            raise RuntimeError("capture-only")
+
+        with patch.object(
+            mcp_server, "verify_preflight_purity", side_effect=_capture
+        ), patch.object(
+            mcp_server, "get_profile", return_value=AUTHOR_PROFILE
+        ), patch.object(
+            mcp_server, "_resolve",
+            return_value=("gitea.prgs.cc", ORG, REPO),
+        ), patch.object(
+            mcp_server, "_auth", return_value="token fake",
+        ), patch.object(
+            mcp_server.role_session_router,
+            "check_author_mutation_after_reviewer_stop",
+            return_value=(True, []),
+        ), patch.object(
+            mcp_server, "_namespace_mutation_block", return_value=None
+        ), patch.object(
+            mcp_server, "_profile_permission_block", return_value=None
+        ):
+            try:
+                mcp_server.gitea_publish_unpublished_issue_branch(**kwargs)
+            except RuntimeError as exc:
+                if "capture-only" not in str(exc) and not captured:
+                    raise
+        self.assertTrue(
+            captured,
+            "gitea_publish_unpublished_issue_branch never called "
+            "verify_preflight_purity",
+        )
+        return captured
+
+    def _base_kwargs(self, **overrides):
+        kwargs = {
+            "issue_number": ISSUE,
+            "branch_name": BRANCH,
+            "worktree_path": "/tmp/issue-815-explicit-worktree",
+            "expected_head": "a" * 40,
+            "remote": "prgs",
+            "org": ORG,
+            "repo": REPO,
+            "git_remote_name": GIT_REMOTE,
+        }
+        kwargs.update(overrides)
+        return kwargs
+
+    def test_explicit_worktree_path_reaches_preflight(self):
+        captured = self._capture_preflight(**self._base_kwargs())
+        self.assertEqual(
+            captured.get("worktree_path"),
+            os.path.realpath(os.path.abspath("/tmp/issue-815-explicit-worktree")),
+            "the authoritative worktree_path must be forwarded into preflight",
+        )
+
+    def test_forwarded_path_is_the_one_publication_uses(self):
+        """AC4 — preflight and publication must judge the same resolved path."""
+        raw = "/tmp/issue-815-explicit-worktree/./"
+        captured = self._capture_preflight(**self._base_kwargs(worktree_path=raw))
+        expected = os.path.realpath(os.path.abspath(raw.strip()))
+        self.assertEqual(captured.get("worktree_path"), expected)
+
+    def test_blank_worktree_path_forwards_none(self):
+        """AC5/AC8 — nothing usable supplied keeps the fail-closed fallback."""
+        for blank in ("", "   "):
+            with self.subTest(blank=repr(blank)):
+                captured = self._capture_preflight(
+                    **self._base_kwargs(worktree_path=blank)
+                )
+                self.assertIsNone(
+                    captured.get("worktree_path"),
+                    "a blank worktree must not resolve to the process cwd",
+                )
+
+    def test_org_repo_and_task_forwarding_are_not_regressed(self):
+        """AC6 — #735's org/repo forwarding and the task name still hold."""
+        captured = self._capture_preflight(**self._base_kwargs())
+        self.assertEqual(captured.get("org"), ORG)
+        self.assertEqual(captured.get("repo"), REPO)
+        self.assertEqual(captured.get("task"), "publish_unpublished_branch")
+
+
+class _ProductionTopologyBase(unittest.TestCase):
+    """Control checkout on master + a distinct registered issue worktree.
+
+    This is the shape the production daemon runs in and the shape the #812
+    fixture never built. ``PROJECT_ROOT`` is the control checkout; the issue
+    worktree is a real registered worktree at a different path; production
+    guards are forced on so the #618 branches-only guard genuinely evaluates.
+    """
+
+    def setUp(self):
+        self.lock_dir = tempfile.TemporaryDirectory()
+        self.addCleanup(self.lock_dir.cleanup)
+        self.origin = tempfile.mkdtemp(prefix="issue815-origin-")
+        self.control = tempfile.mkdtemp(prefix="issue815-control-")
+        for path in (self.origin, self.control):
+            self.addCleanup(
+                lambda p=path: subprocess.run(["rm", "-rf", p], check=False)
+            )
+        self._init_repos()
+        self.remotes = patch.dict(
+            mcp_server.REMOTES,
+            {"prgs": {"host": "gitea.prgs.cc", "org": ORG, "repo": REPO}},
+        )
+        self.remotes.start()
+        self.addCleanup(patch.stopall)
+        mcp_server._IDENTITY_CACHE.clear()
+
+    def _git(self, *args, cwd=None):
+        return subprocess.run(
+            ["git", "-C", cwd or self.control, *args],
+            capture_output=True, text=True, check=True,
+        )
+
+    def _init_repos(self):
+        subprocess.run(
+            ["git", "init", "-q", "--bare", "-b", "master", self.origin], check=True
+        )
+        self._git("init", "-q", "-b", "master")
+        self._git("config", "user.email", "[email protected]")
+        self._git("config", "user.name", "Test")
+        self._git("remote", "add", GIT_REMOTE, self.origin)
+
+        with open(os.path.join(self.control, "seed.txt"), "w") as fh:
+            fh.write("seed\n")
+        # The real repository gitignores branches/, so a registered worktree
+        # living there does not dirty the stable control checkout. Mirror that,
+        # or the #615 dirty-runtime block fires on the worktree we just created.
+        with open(os.path.join(self.control, ".gitignore"), "w") as fh:
+            fh.write("branches/\n")
+        self._git("add", "seed.txt", ".gitignore")
+        self._git("commit", "-q", "-m", "seed")
+        self.base_sha = self._git("rev-parse", "HEAD").stdout.strip()
+        self._git("push", "-q", GIT_REMOTE, "master")
+
+        # The control checkout STAYS on master. This is the whole point: the
+        # daemon's process root is the stable control checkout, never the
+        # worktree the publication targets.
+        self.worktree = os.path.realpath(
+            os.path.join(self.control, "branches", WORKTREE_DIRNAME)
+        )
+        self._git("worktree", "add", "-q", "-b", BRANCH, self.worktree, "master")
+
+        with open(os.path.join(self.worktree, "work.txt"), "w") as fh:
+            fh.write("unpublished implementation\n")
+        self._git("add", "work.txt", cwd=self.worktree)
+        self._git("commit", "-q", "-m", "unpublished implementation", cwd=self.worktree)
+        self.head_sha = self._git("rev-parse", "HEAD", cwd=self.worktree).stdout.strip()
+
+        self.control_branch = self._git(
+            "rev-parse", "--abbrev-ref", "HEAD"
+        ).stdout.strip()
+
+    # ── durable lock naming the caller and the issue worktree ────────────
+    def lock_path(self):
+        return issue_lock_store.lock_file_path(
+            remote="prgs", org=ORG, repo=REPO, issue_number=ISSUE,
+            lock_dir=self.lock_dir.name,
+        )
+
+    def write_lock(self, *, bind_session=True, **overrides):
+        path = self.lock_path()
+        claimant = overrides.pop(
+            "claimant", {"username": IDENTITY, "profile": PROFILE}
+        )
+        pid = overrides.pop("session_pid", os.getpid())
+        lease = {
+            "operation_type": issue_lock_store.AUTHOR_ISSUE_WORK_LEASE,
+            "issue_number": ISSUE,
+            "pr_number": None,
+            "branch": overrides.get("branch_name", BRANCH),
+            "worktree_path": overrides.get("worktree_path", self.worktree),
+            "claimant": claimant,
+            "created_at": _ts(-2),
+            "last_heartbeat_at": _ts(-2),
+            "expires_at": _ts(-1),
+        }
+        lease.update(overrides.pop("work_lease", {}))
+        data = {
+            "issue_number": ISSUE,
+            "branch_name": BRANCH,
+            "remote": "prgs",
+            "org": ORG,
+            "repo": REPO,
+            "worktree_path": self.worktree,
+            "session_pid": pid,
+            "pid": pid,
+            "lock_generation": 1,
+            "work_lease": lease,
+            "lock_provenance": issue_lock_provenance.build_sanctioned_lock_provenance(
+                tool="gitea_lock_issue", claimant=claimant
+            ),
+        }
+        data.update(overrides)
+        data["lock_file_path"] = path
+        issue_lock_store.save_lock_file(path, data)
+        # Bind the session pointer so the #683 issue-scope guard resolves an
+        # owning issue for this author session. In real production the publish
+        # task does not require a session lock — require_author_lock is keyed on
+        # the test-only production_guards_forced() flag, which this suite must
+        # set to make preflight run at all — so this pointer is fixture
+        # scaffolding to clear a guard production would not apply here, never a
+        # softening of the worktree-forwarding behaviour under test. The
+        # preflight-negative cases below leave it unbound precisely so the #618
+        # guard is reached with no session fallback to rescue a bad worktree.
+        if bind_session:
+            pointer = {
+                "pid": os.getpid(),
+                "lock_file_path": path,
+                "issue_number": ISSUE,
+                "branch_name": data["branch_name"],
+                "remote": "prgs",
+                "org": ORG,
+                "repo": REPO,
+            }
+            issue_lock_store.save_lock_file(
+                issue_lock_store.session_pointer_path(self.lock_dir.name), pointer
+            )
+        return path
+
+    def _tool_env(self):
+        env = shared_mutation_env(
+            PROFILE, include_example_repo=True,
+            GITEA_ISSUE_LOCK_DIR=self.lock_dir.name,
+        )
+        env["GITEA_ISSUE_LOCK_DIR"] = self.lock_dir.name
+        # The defect only exists where preflight actually runs. Under pytest the
+        # production root/branches/scope guards are skipped unless forced on, so
+        # force them: this test exists to exercise the #618 guard, not to bypass
+        # it. Parity is pinned to the server's own startup head so repointing
+        # PROJECT_ROOT does not read as a stale daemon.
+        env["GITEA_TEST_FORCE_PRODUCTION_GUARDS"] = "1"
+        # Production is a promoted stable-control runtime. The pytest process
+        # itself runs from a branches/ worktree, which the #615 runtime-mode
+        # gate correctly classifies as dev-test; declaring the sanctioned mode
+        # models the production daemon rather than defeating the gate. Without
+        # this, forcing production guards on would trip the *runtime-mode* block
+        # for a reason unrelated to the #815 worktree-forwarding defect.
+        env["GITEA_MCP_RUNTIME_MODE"] = "stable-control"
+        startup_head = mcp_server._STARTUP_PARITY.get("startup_head") or ""
+        env["GITEA_TEST_CURRENT_HEAD"] = startup_head
+        env["GITEA_TEST_LIVE_REMOTE_HEAD"] = startup_head
+        return env
+
+    def run_publish(self, *, open_prs=None, expected_head=None, **kwargs):
+        """Drive the public tool with PROJECT_ROOT pinned to the CONTROL checkout."""
+        env = self._tool_env()
+        with patch(
+            "mcp_server._list_open_pulls", return_value=list(open_prs or [])
+        ), patch(
+            "mcp_server._auth", return_value="token x"
+        ), patch(
+            "mcp_server.get_auth_header", return_value="token x"
+        ), patch(
+            "mcp_server._work_lease_claimant",
+            return_value={"username": IDENTITY, "profile": PROFILE},
+        ), patch.object(
+            # NOTE: the control checkout — deliberately NOT self.worktree.
+            mcp_server, "PROJECT_ROOT", self.control
+        ), patch.dict(os.environ, env, clear=True):
+            os.environ["GITEA_ISSUE_LOCK_DIR"] = self.lock_dir.name
+            return mcp_server.gitea_publish_unpublished_issue_branch(
+                issue_number=kwargs.pop("issue_number", ISSUE),
+                branch_name=kwargs.pop("branch_name", BRANCH),
+                worktree_path=kwargs.pop("worktree_path", self.worktree),
+                expected_head=expected_head or self.head_sha,
+                remote="prgs",
+                git_remote_name=kwargs.pop("git_remote_name", GIT_REMOTE),
+                **kwargs,
+            )
+
+    def remote_head(self, branch=BRANCH):
+        res = subprocess.run(
+            ["git", "-C", self.origin, "rev-parse", "--verify", "--quiet", branch],
+            capture_output=True, text=True, check=False,
+        )
+        return (res.stdout or "").strip() or None
+
+
+class TestForwardingClearsThe618Guard(_ProductionTopologyBase):
+    """AC2 — the faithful production reproduction, and the sharpest fix proof.
+
+    The production recovery worker had **no** session issue lock — acquiring one
+    was the very thing the deadlock prevented — so preflight had nothing but the
+    explicit ``worktree_path`` argument to resolve the workspace from. This class
+    reproduces exactly that: no session pointer is bound, so there is no
+    author-lock fallback to rescue a dropped argument.
+
+    With the argument forwarded (fixed source) the #618 branches-only guard
+    accepts the registered issue worktree and the call advances to the next
+    guard. With the argument dropped (the buggy source this issue reports)
+    preflight falls back to ``PROJECT_ROOT`` — the stable control checkout — and
+    the #618 guard traps the call there. The two outcomes are told apart by the
+    guard that fired, on its own error text.
+
+    This test therefore *fails* against the unpatched source (the call is trapped
+    at #618 instead of clearing it), which is what makes it a regression rather
+    than a smoke test.
+    """
+
+    _CONTROL_CHECKOUT_MARKERS = ("stable control checkout", "#618")
+
+    def test_explicit_worktree_clears_618_without_a_session_lock(self):
+        # No write_lock(): the session is deliberately unbound, as in production.
+        with self.assertRaises(RuntimeError) as ctx:
+            self.run_publish()
+        message = str(ctx.exception)
+        # The workspace guard is satisfied — the failure is the *later* scope
+        # guard (no owning issue), never the control-checkout refusal. If the
+        # argument were dropped, this call would be trapped at #618 instead.
+        for marker in self._CONTROL_CHECKOUT_MARKERS:
+            self.assertNotIn(
+                marker, message,
+                f"the explicit worktree must clear #618; got a control-checkout "
+                f"refusal instead: {message}",
+            )
+        self.assertIn(
+            "owning issue", message,
+            f"expected the downstream scope guard to fire, got: {message}",
+        )
+        self.assertIsNone(self.remote_head())
+
+    def test_dropped_argument_would_be_trapped_at_618(self):
+        # Simulate the buggy call shape directly: no session lock, and preflight
+        # given no worktree, exactly as the unpatched source left it. This pins
+        # the control-checkout refusal that the fix eliminates, so the pair of
+        # tests brackets the defect from both sides regardless of which source
+        # version is loaded.
+        env = self._tool_env()
+        with patch(
+            "mcp_server._list_open_pulls", return_value=[]
+        ), patch(
+            "mcp_server._auth", return_value="token x"
+        ), patch(
+            "mcp_server.get_auth_header", return_value="token x"
+        ), patch(
+            "mcp_server._work_lease_claimant",
+            return_value={"username": IDENTITY, "profile": PROFILE},
+        ), patch.object(
+            mcp_server, "PROJECT_ROOT", self.control
+        ), patch.dict(os.environ, env, clear=True):
+            os.environ["GITEA_ISSUE_LOCK_DIR"] = self.lock_dir.name
+            with self.assertRaises(RuntimeError) as ctx:
+                # Drive verify_preflight_purity the way the buggy body did:
+                # no worktree_path forwarded at all.
+                mcp_server.verify_preflight_purity(
+                    "prgs",
+                    task="publish_unpublished_branch",
+                    org=ORG,
+                    repo=REPO,
+                )
+        message = str(ctx.exception)
+        self.assertTrue(
+            any(m in message for m in self._CONTROL_CHECKOUT_MARKERS),
+            f"a dropped worktree must trap at the control checkout: {message}",
+        )
+        self.assertIsNone(self.remote_head())
+
+
+class TestProductionTopologyPublishes(_ProductionTopologyBase):
+    """AC2/AC4/AC7 — the explicit registered worktree is what preflight validates."""
+
+    def test_fixture_is_genuinely_the_production_topology(self):
+        """Guard the guard: if this drifts, the regression stops meaning anything."""
+        self.assertNotEqual(
+            os.path.realpath(self.control), self.worktree,
+            "the issue worktree must not be PROJECT_ROOT",
+        )
+        self.assertEqual(
+            self.control_branch, "master",
+            "the control checkout must sit on a stable branch",
+        )
+        self.assertTrue(
+            os.path.realpath(self.worktree).startswith(
+                os.path.realpath(os.path.join(self.control, "branches")) + os.sep
+            ),
+            "the issue worktree must live under branches/",
+        )
+        listed = subprocess.run(
+            ["git", "-C", self.control, "worktree", "list"],
+            capture_output=True, text=True, check=True,
+        ).stdout
+        self.assertIn(
+            self.worktree, listed, "the issue worktree must be genuinely registered"
+        )
+
+    def test_publishes_from_a_control_rooted_daemon(self):
+        """The exact production failure: this refused with #618 before the fix."""
+        self.write_lock()
+        self.assertIsNone(self.remote_head(), "fixture must start unpublished")
+        res = self.run_publish()
+        self.assertTrue(res.get("success"), res)
+        self.assertTrue(res.get("performed"), res)
+        self.assertEqual(self.remote_head(), self.head_sha)
+
+    def test_dry_run_uses_the_explicit_worktree(self):
+        """AC4 — dry-run reaches the same decision without publishing."""
+        self.write_lock()
+        res = self.run_publish(dry_run=True)
+        self.assertTrue(res.get("success"), res)
+        self.assertFalse(res.get("performed"), res)
+        self.assertTrue(res.get("would_publish"), res)
+        self.assertIsNone(self.remote_head(), "dry-run must not publish")
+
+    def test_dry_run_and_apply_agree_on_the_same_worktree(self):
+        """AC4 — both paths resolve the same workspace, so both succeed."""
+        self.write_lock()
+        dry = self.run_publish(dry_run=True)
+        self.assertTrue(dry.get("would_publish"), dry)
+        applied = self.run_publish()
+        self.assertTrue(applied.get("performed"), applied)
+        self.assertEqual(self.remote_head(), self.head_sha)
+
+    def test_read_after_write_verification_still_runs(self):
+        """AC6 — PR #814's post-publication verification is unchanged."""
+        self.write_lock()
+        res = self.run_publish()
+        self.assertTrue(res.get("verified"), res)
+        self.assertEqual(res.get("remote_head_sha"), self.head_sha)
+
+
+class TestProductionTopologyFailsClosed(_ProductionTopologyBase):
+    """AC3/AC5/AC8 — the fix does not weaken any refusal.
+
+    A refusal reaches the caller by one of two mechanisms, and this class holds
+    them apart deliberately. A bad *workspace* is caught by the #618 preflight
+    guard, which raises before the assessor is built. A bad *content/ownership*
+    fact passes preflight (the worktree itself is fine) and is then refused by
+    the publication assessor, which returns ``success: False``. Both are
+    fail-closed; asserting the wrong mechanism would hide a regression.
+    """
+
+    # ── #618 preflight refusals: no session lock, so nothing rescues a bad
+    #    workspace and the guard fires exactly as it does in production ──────
+    def _assert_preflight_raises(self, **kwargs):
+        with self.assertRaises(RuntimeError) as ctx:
+            self.run_publish(**kwargs)
+        self.assertIsNone(
+            self.remote_head(), "a blocked publication must not reach the remote"
+        )
+        return str(ctx.exception)
+
+    def test_blank_worktree_path_fails_closed_via_618(self):
+        """AC5 — a blank path forwards None, so preflight sees the control root."""
+        for blank in ("", "   "):
+            with self.subTest(blank=repr(blank)):
+                message = self._assert_preflight_raises(worktree_path=blank)
+                self.assertIn("618", message)
+
+    def test_control_checkout_as_worktree_fails_closed_via_618(self):
+        """AC5 — naming the stable control checkout explicitly is still refused."""
+        message = self._assert_preflight_raises(worktree_path=self.control)
+        self.assertIn("618", message)
+
+    def test_unregistered_directory_fails_closed(self):
+        """AC3 — a plain directory under branches/ is not a registered worktree."""
+        bogus = os.path.join(self.control, "branches", "not-a-worktree")
+        os.makedirs(bogus, exist_ok=True)
+        self._assert_preflight_raises(worktree_path=bogus)
+
+    def test_missing_worktree_path_fails_closed(self):
+        """AC3 — a path that does not exist is refused, not silently replaced."""
+        missing = os.path.join(self.control, "branches", "absent-worktree")
+        self._assert_preflight_raises(worktree_path=missing)
+
+    # ── assessor refusals: preflight passes on a valid worktree, then the
+    #    publication assessor refuses on content/ownership evidence ──────────
+    def _assert_assessor_refuses(self, **kwargs):
+        res = self.run_publish(**kwargs)
+        self.assertFalse(res.get("success"), res)
+        self.assertFalse(res.get("performed"), res)
+        self.assertIsNone(self.remote_head())
+        return res
+
+    def test_changed_local_head_still_refuses(self):
+        """AC6 — the declared expected_head remains authoritative."""
+        self.write_lock()
+        self._assert_assessor_refuses(expected_head="b" * 40)
+
+    def test_foreign_claimant_still_refuses(self):
+        """AC6 — ownership still comes from the durable lock record."""
+        self.write_lock(claimant={"username": "someone-else", "profile": PROFILE})
+        self._assert_assessor_refuses()
+
+    def test_dirty_worktree_still_refuses(self):
+        """AC6 — cleanliness enforcement survives the forwarding change."""
+        self.write_lock()
+        with open(os.path.join(self.worktree, "work.txt"), "a") as fh:
+            fh.write("uncommitted drift\n")
+        self._assert_assessor_refuses()
+
+    def test_competing_open_pr_still_refuses(self):
+        """AC6 — a rival claim on another branch still blocks."""
+        self.write_lock()
+        self._assert_assessor_refuses(
+            open_prs=[{"number": 4242, "head": {"ref": f"fix/issue-{ISSUE}-rival"}}]
+        )
+
+    def test_issue_lock_record_is_not_mutated_by_a_refusal(self):
+        """AC6 — record separation (#812 AC23) is unaffected by this change."""
+        path = self.write_lock()
+        with open(path, "rb") as fh:
+            before = fh.read()
+        self._assert_assessor_refuses(expected_head="c" * 40)
+        with open(path, "rb") as fh:
+            self.assertEqual(before, fh.read())
+
+
+class TestProtectedFixtureNotReferenced(unittest.TestCase):
+    """AC9 — this regression never names the protected #635 fixture.
+
+    The forbidden tokens are reconstructed from fragments so this assertion
+    file does not itself contain them and produce a false positive.
+    """
+
+    def test_no_reference_to_the_protected_worktree(self):
+        forbidden = [
+            "issue-635-" + "project-registry-api",
+            "b2f6e9a6dc40e9651ef8" + "76f322dd0a68bddebfd8",
+        ]
+        here = os.path.abspath(__file__)
+        with open(here, "r", encoding="utf-8") as fh:
+            text = fh.read()
+        for token in forbidden:
+            self.assertNotIn(
+                token, text,
+                f"the protected #635 fixture must not be referenced: {token}",
+            )
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_master_parity_gate.py

@@ -78,6 +78,95 @@ class TestBlockReasonsAndReport(unittest.TestCase):
         self.assertTrue(report["recovery"])
 
 
+class TestLiveRemoteParity(unittest.TestCase):
+    """#610: parity must account for the live remote master, not just local.
+
+    The daemon can be stale relative to the live remote target while the local
+    checkout HEAD still matches the daemon's startup commit, so local parity
+    reports green even though a mutation would run against outdated code.
+    """
+
+    SHA_C = "c" * 40
+
+    def test_distinguishes_three_shas(self):
+        res = mp.assess_master_parity(
+            {"startup_head": SHA_A}, SHA_A, live_remote_head=SHA_B)
+        self.assertEqual(res["daemon_start_head"], SHA_A)
+        self.assertEqual(res["local_head"], SHA_A)
+        self.assertEqual(res["live_remote_head"], SHA_B)
+
+    def test_mutation_safe_only_when_all_three_match(self):
+        res = mp.assess_master_parity(
+            {"startup_head": SHA_A}, SHA_A, live_remote_head=SHA_A)
+        self.assertTrue(res["mutation_safe"])
+        self.assertTrue(res["live_known"])
+        self.assertFalse(res["live_stale"])
+
+    def test_live_stale_when_remote_advanced_past_daemon(self):
+        # Local checkout still matches the daemon start (local parity green),
+        # but the live remote master has advanced -> daemon is live-stale.
+        res = mp.assess_master_parity(
+            {"startup_head": SHA_A}, SHA_A, live_remote_head=SHA_B)
+        self.assertTrue(res["in_parity"])  # local parity still green
+        self.assertTrue(res["live_stale"])
+        self.assertFalse(res["mutation_safe"])
+        self.assertTrue(any("live" in r.lower() for r in res["reasons"]))
+
+    def test_live_unknown_is_not_mutation_safe_but_not_stale(self):
+        # Non-goal: unfetchable live remote must not be treated as stale for
+        # read-only, but a mutation-safe claim fails closed.
+        res = mp.assess_master_parity(
+            {"startup_head": SHA_A}, SHA_A, live_remote_head=None)
+        self.assertFalse(res["live_known"])
+        self.assertFalse(res["mutation_safe"])
+        self.assertFalse(res["live_stale"])
+        self.assertTrue(res["in_parity"])
+
+    def test_default_live_remote_preserves_legacy_shape(self):
+        # Callers that do not supply a live head keep the pre-#610 behavior:
+        # in-parity, not live-stale, no live-derived block.
+        res = mp.assess_master_parity({"startup_head": SHA_A}, SHA_A)
+        self.assertFalse(res["live_stale"])
+        self.assertEqual(mp.parity_block_reasons(res), [])
+
+
+class TestLiveStaleBlockAndReport(unittest.TestCase):
+    """#610: live-staleness must block mutations and surface a typed blocker."""
+
+    def test_live_stale_produces_block_reasons(self):
+        res = mp.assess_master_parity(
+            {"startup_head": SHA_A}, SHA_A, live_remote_head=SHA_B)
+        self.assertTrue(mp.parity_block_reasons(res))
+
+    def test_disable_env_suppresses_live_stale_block(self):
+        res = mp.assess_master_parity(
+            {"startup_head": SHA_A}, SHA_A, live_remote_head=SHA_B)
+        with patch.dict(os.environ, {mp.ENV_DISABLE: "1"}):
+            self.assertEqual(mp.parity_block_reasons(res), [])
+
+    def test_resolver_disagreement_returns_typed_blocker(self):
+        # Parity says local-green, resolver says restart required -> disagreement
+        # is a typed, fail-closed blocker naming the resolver as authoritative.
+        res = mp.assess_master_parity({"startup_head": SHA_A}, SHA_A)
+        blocker = mp.parity_resolver_disagreement(res, resolver_restart_required=True)
+        self.assertIsNotNone(blocker)
+        self.assertEqual(blocker["kind"], "parity_resolver_disagreement")
+        self.assertTrue(blocker["restart_required"])
+        self.assertTrue(blocker["resolver_authoritative"])
+
+    def test_no_disagreement_when_resolver_agrees(self):
+        res = mp.assess_master_parity({"startup_head": SHA_A}, SHA_A)
+        self.assertIsNone(
+            mp.parity_resolver_disagreement(res, resolver_restart_required=False))
+
+    def test_live_stale_report_names_live_remote(self):
+        res = mp.assess_master_parity(
+            {"startup_head": SHA_A}, SHA_A, live_remote_head=SHA_B)
+        report = mp.parity_report(res)
+        self.assertEqual(report["live_remote_head"], SHA_B)
+        self.assertTrue(report["restart_required"])
+
+
 class TestReadGitHead(unittest.TestCase):
     def test_test_override_takes_precedence(self):
         with patch.dict(os.environ, {mp.ENV_TEST_CURRENT_HEAD: SHA_B}):
@@ -95,6 +184,149 @@ class TestReadGitHead(unittest.TestCase):
             self.assertIsNone(mp.read_git_head(""))
 
 
+class TestReadRemoteMasterHead(unittest.TestCase):
+    """#610: live remote master head reader (env-overridable, fails to None)."""
+
+    def test_test_override_takes_precedence(self):
+        with patch.dict(os.environ, {mp.ENV_TEST_LIVE_REMOTE_HEAD: SHA_B}):
+            self.assertEqual(mp.read_remote_master_head("/nonexistent"), SHA_B)
+
+    def test_blank_override_is_none(self):
+        with patch.dict(os.environ, {mp.ENV_TEST_LIVE_REMOTE_HEAD: "   "}):
+            self.assertIsNone(mp.read_remote_master_head("/nonexistent"))
+
+    def test_unfetchable_remote_is_none(self):
+        # No override; a bogus root/remote must fail closed to None, never raise.
+        env = {k: v for k, v in os.environ.items()
+               if k != mp.ENV_TEST_LIVE_REMOTE_HEAD}
+        with patch.dict(os.environ, env, clear=True):
+            self.assertIsNone(
+                mp.read_remote_master_head("/nonexistent", remote="nope"))
+
+
+class TestRemoteHeadCache(unittest.TestCase):
+    """#610: live remote reads are cached with a TTL to stay off the network.
+
+    The parity gate runs on every mutation and every runtime-context read, so an
+    unbounded ``git ls-remote`` per call would be a latency/flakiness regression.
+    """
+
+    def setUp(self):
+        # These cases intentionally exercise the subprocess/cache path, so they
+        # opt out of suite-wide hermetic mode (PR #788 F1).
+        self._saved_hermetic = mp.hermetic_test_mode()
+        mp.set_hermetic_test_mode(False)
+        mp._clear_remote_head_cache()
+        env = {
+            k: v for k, v in os.environ.items()
+            if k not in (mp.ENV_TEST_LIVE_REMOTE_HEAD,
+                         mp.ENV_TEST_ALLOW_LIVE_REMOTE_PROBE,
+                         "PYTEST_CURRENT_TEST")
+        }
+        # Allow the probe path under hermetic defenses while still mocking
+        # subprocess so no real network call runs.
+        env[mp.ENV_TEST_ALLOW_LIVE_REMOTE_PROBE] = "1"
+        self._env = patch.dict(os.environ, env, clear=True)
+        self._env.start()
+        self.addCleanup(self._env.stop)
+        self.addCleanup(mp._clear_remote_head_cache)
+        self.addCleanup(
+            lambda: mp.set_hermetic_test_mode(self._saved_hermetic)
+        )
+
+    def _fake_run(self, sha):
+        class _R:
+            returncode = 0
+            stdout = f"{sha}\trefs/heads/master\n"
+        calls = {"n": 0}
+
+        def run(*args, **kwargs):
+            calls["n"] += 1
+            return _R()
+        return run, calls
+
+    def test_second_call_within_ttl_uses_cache(self):
+        run, calls = self._fake_run(SHA_B)
+        with patch.object(mp.subprocess, "run", run):
+            a = mp.read_remote_master_head("/repo", remote="prgs", ttl=100)
+            b = mp.read_remote_master_head("/repo", remote="prgs", ttl=100)
+        self.assertEqual(a, SHA_B)
+        self.assertEqual(b, SHA_B)
+        self.assertEqual(calls["n"], 1)
+
+    def test_zero_ttl_bypasses_cache(self):
+        run, calls = self._fake_run(SHA_B)
+        with patch.object(mp.subprocess, "run", run):
+            mp.read_remote_master_head("/repo", remote="prgs", ttl=0)
+            mp.read_remote_master_head("/repo", remote="prgs", ttl=0)
+        self.assertEqual(calls["n"], 2)
+
+    def test_env_override_never_touches_subprocess(self):
+        run, calls = self._fake_run(SHA_B)
+        with patch.dict(os.environ, {mp.ENV_TEST_LIVE_REMOTE_HEAD: SHA_A}):
+            with patch.object(mp.subprocess, "run", run):
+                self.assertEqual(
+                    mp.read_remote_master_head("/repo", remote="prgs"), SHA_A)
+        self.assertEqual(calls["n"], 0)
+
+
+class TestHermeticLiveRemoteReads(unittest.TestCase):
+    """#610 / PR #788 F1/F2: suite hermetic mode never hits the network."""
+
+    def setUp(self):
+        self._saved = mp.hermetic_test_mode()
+        mp.set_hermetic_test_mode(True)
+        mp._clear_remote_head_cache()
+        self.addCleanup(lambda: mp.set_hermetic_test_mode(self._saved))
+        self.addCleanup(mp._clear_remote_head_cache)
+
+    def test_hermetic_mode_returns_none_without_subprocess(self):
+        run_calls = {"n": 0}
+
+        def boom(*args, **kwargs):
+            run_calls["n"] += 1
+            raise AssertionError("ls-remote must not run under hermetic mode")
+
+        env = {
+            k: v for k, v in os.environ.items()
+            if k not in (mp.ENV_TEST_LIVE_REMOTE_HEAD,
+                         mp.ENV_TEST_ALLOW_LIVE_REMOTE_PROBE)
+        }
+        with patch.dict(os.environ, env, clear=True):
+            with patch.object(mp.subprocess, "run", boom):
+                self.assertIsNone(
+                    mp.read_remote_master_head("/repo", remote="prgs")
+                )
+        self.assertEqual(run_calls["n"], 0)
+
+    def test_hermetic_mode_survives_clear_true_env(self):
+        """Module flag, not env pin: clear=True cannot re-enable the probe."""
+        run_calls = {"n": 0}
+
+        def boom(*args, **kwargs):
+            run_calls["n"] += 1
+            raise AssertionError("ls-remote must not run after clear=True")
+
+        with patch.dict(os.environ, {}, clear=True):
+            with patch.object(mp.subprocess, "run", boom):
+                self.assertIsNone(mp.read_remote_master_head("/repo"))
+        self.assertEqual(run_calls["n"], 0)
+
+    def test_explicit_override_still_wins_under_hermetic(self):
+        run_calls = {"n": 0}
+
+        def boom(*args, **kwargs):
+            run_calls["n"] += 1
+            raise AssertionError("override must bypass subprocess")
+
+        with patch.dict(os.environ, {mp.ENV_TEST_LIVE_REMOTE_HEAD: SHA_B}):
+            with patch.object(mp.subprocess, "run", boom):
+                self.assertEqual(
+                    mp.read_remote_master_head("/repo"), SHA_B
+                )
+        self.assertEqual(run_calls["n"], 0)
+
+
 class TestServerWiring(unittest.TestCase):
     """Integration with the gate choke point in the server namespace."""
 
@@ -105,6 +337,13 @@ class TestServerWiring(unittest.TestCase):
         self._saved = self.srv._STARTUP_PARITY
         self.srv._STARTUP_PARITY = {"root": self.srv.PROJECT_ROOT,
                                     "startup_head": SHA_A}
+        # Keep the live-remote read hermetic (no real ls-remote network call):
+        # default the live master to the daemon start so parity is fully green
+        # unless a test overrides the live head explicitly (#610).
+        self._live_patch = patch.dict(
+            os.environ, {mp.ENV_TEST_LIVE_REMOTE_HEAD: SHA_A})
+        self._live_patch.start()
+        self.addCleanup(self._live_patch.stop)
 
     def tearDown(self):
         self.srv._STARTUP_PARITY = self._saved
@@ -147,6 +386,36 @@ class TestServerWiring(unittest.TestCase):
             self.assertTrue(out["in_parity"])
             self.assertNotIn("report", out)
 
+    # --- #610: live-remote wiring -------------------------------------------
+
+    def test_live_stale_blocks_mutation_though_local_green(self):
+        # Local checkout matches the daemon start (local parity green) but the
+        # live remote master has advanced -> mutations must fail closed.
+        with patch.dict(os.environ, {mp.ENV_TEST_CURRENT_HEAD: SHA_A,
+                                     mp.ENV_TEST_LIVE_REMOTE_HEAD: SHA_B}):
+            self.assertEqual(self.srv._master_parity_block("gitea.read"), [])
+            self.assertTrue(
+                self.srv._master_parity_block("gitea.pr.create"))
+
+    def test_assess_tool_exposes_three_distinct_shas(self):
+        with patch.dict(os.environ, {mp.ENV_TEST_CURRENT_HEAD: SHA_A,
+                                     mp.ENV_TEST_LIVE_REMOTE_HEAD: SHA_B}):
+            out = self.srv.gitea_assess_master_parity(remote="prgs")
+            self.assertEqual(out["daemon_start_head"], SHA_A)
+            self.assertEqual(out["local_head"], SHA_A)
+            self.assertEqual(out["live_remote_head"], SHA_B)
+            self.assertTrue(out["live_stale"])
+            self.assertFalse(out["mutation_safe"])
+            self.assertIn("report", out)
+
+    def test_assess_tool_mutation_safe_when_all_three_match(self):
+        with patch.dict(os.environ, {mp.ENV_TEST_CURRENT_HEAD: SHA_A,
+                                     mp.ENV_TEST_LIVE_REMOTE_HEAD: SHA_A}):
+            out = self.srv.gitea_assess_master_parity(remote="prgs")
+            self.assertTrue(out["mutation_safe"])
+            self.assertFalse(out["live_stale"])
+            self.assertNotIn("report", out)
+
 
 if __name__ == "__main__":
     unittest.main()

tests/test_task_capability_role_invariants.py

@@ -139,6 +139,9 @@ EXPECTED_ROLE_EXCLUSIVE_TASKS = frozenset(
         "gitea_release_merger_pr_lease",
         "create_branch",
         "push_branch",
+        # #812 AC20: publishing an unpublished local head is author-only for the
+        # same reason every other push is — it writes a branch to the remote.
+        "publish_unpublished_branch",
         "create_pr",
         "commit_files",
         "gitea_commit_files",

tests/test_webui_worker_registry.py

@@ -0,0 +1,458 @@
+"""Tests for the worker registry and configuration schema (#798, epic #797)."""
+import json
+import sys
+import tempfile
+import unittest
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+
+from webui.worker_registry import (
+    ALLOWED_ROLES,
+    SCHEMA_VERSION,
+    RegistryValidationError,
+    WorkerRegistry,
+    default_registry_path,
+    find_provider,
+    find_worker,
+    history_dir,
+    list_revisions,
+    load_registry,
+    registry_to_dict,
+    registry_to_document,
+    rollback_to_revision,
+    save_registry,
+    validate_payload,
+    worker_to_dict,
+    workers_for_provider,
+)
+
+_EXPECTED_PROVIDER_IDS = ("claude", "grok", "codex", "agy", "kimi-k")
+
+
+def _provider(provider_id: str = "claude", **overrides) -> dict:
+    payload = {
+        "id": provider_id,
+        "display_name": "Claude",
+        "vendor": "Anthropic",
+        "executable": "claude",
+        "available": True,
+        "models": ["claude-opus-4-8"],
+        "notes": "",
+    }
+    payload.update(overrides)
+    return payload
+
+
+def _worker(worker_id: str = "claude-author", **overrides) -> dict:
+    payload = {
+        "id": worker_id,
+        "display_name": "Claude author",
+        "provider": "claude",
+        "model": "claude-opus-4-8",
+        "project": "gitea-tools",
+        "role": "author",
+        "namespace": "gitea-author",
+        "profile": "prgs-author",
+        "workflow": "skills/llm-project-workflow/workflows/work-issue.md",
+        "schedule": {"kind": "cron", "expression": "0 * * * *"},
+        "timeout_seconds": 3600,
+        "enabled": True,
+        "scheduler": {"kind": "launchd", "label": "cc.prgs.claude.author"},
+        "notes": "",
+    }
+    payload.update(overrides)
+    return payload
+
+
+def _document(providers=None, workers=None, **overrides) -> dict:
+    payload = {
+        "version": SCHEMA_VERSION,
+        "revision": 1,
+        "updated_at": "2026-07-22T00:00:00Z",
+        "providers": providers if providers is not None else [_provider()],
+        "workers": workers if workers is not None else [_worker()],
+    }
+    payload.update(overrides)
+    return payload
+
+
+class _TempRegistryCase(unittest.TestCase):
+    """Base case giving each test an isolated registry file."""
+
+    def setUp(self):
+        self._tmp = tempfile.TemporaryDirectory()
+        self.addCleanup(self._tmp.cleanup)
+        self.path = Path(self._tmp.name) / "workers.registry.json"
+
+    def write(self, document: dict) -> Path:
+        self.path.write_text(json.dumps(document, indent=2) + "\n", encoding="utf-8")
+        return self.path
+
+    def parse(self, document: dict) -> WorkerRegistry:
+        return validate_payload(document, source_path=self.path)
+
+
+class TestPackagedRegistry(unittest.TestCase):
+    """AC: the declarative registry is the source of truth and ships with the app."""
+
+    def test_default_path_points_at_packaged_data(self):
+        path = default_registry_path()
+        self.assertEqual(path.name, "workers.registry.json")
+        self.assertEqual(path.parent.name, "data")
+
+    def test_packaged_registry_loads_and_validates(self):
+        registry = load_registry()
+        self.assertEqual(registry.version, SCHEMA_VERSION)
+        self.assertGreaterEqual(registry.revision, 1)
+
+    def test_packaged_registry_declares_all_five_providers(self):
+        registry = load_registry()
+        self.assertEqual(
+            tuple(provider.id for provider in registry.providers),
+            _EXPECTED_PROVIDER_IDS,
+        )
+
+    def test_packaged_registry_carries_no_credentials(self):
+        raw = default_registry_path().read_text(encoding="utf-8").lower()
+        for marker in ("token", "password", "secret", "api_key", "credential"):
+            self.assertNotIn(marker, raw)
+
+
+class TestSeparateEntities(_TempRegistryCase):
+    """AC: providers and configured workers are separate entities."""
+
+    def test_provider_may_exist_with_no_workers(self):
+        registry = self.parse(
+            _document(providers=[_provider("grok", display_name="Grok")], workers=[])
+        )
+        self.assertEqual(len(registry.providers), 1)
+        self.assertEqual(registry.workers, ())
+        self.assertEqual(workers_for_provider(registry, "grok"), ())
+
+    def test_many_workers_may_share_one_provider(self):
+        registry = self.parse(
+            _document(
+                workers=[
+                    _worker("claude-author"),
+                    _worker(
+                        "claude-reviewer",
+                        role="reviewer",
+                        namespace="gitea-reviewer",
+                        profile="prgs-reviewer",
+                        scheduler={"kind": "launchd", "label": "cc.prgs.claude.reviewer"},
+                    ),
+                ]
+            )
+        )
+        self.assertEqual(len(workers_for_provider(registry, "claude")), 2)
+        self.assertEqual(len(registry.providers), 1)
+
+    def test_worker_referencing_unknown_provider_is_refused(self):
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(workers=[_worker(provider="mystery")]))
+        self.assertIn("unknown provider", str(ctx.exception))
+
+    def test_lookup_helpers(self):
+        registry = self.parse(_document())
+        self.assertIsNotNone(find_worker(registry, "claude-author"))
+        self.assertIsNone(find_worker(registry, "absent"))
+        self.assertIsNotNone(find_provider(registry, "claude"))
+        self.assertIsNone(find_provider(registry, "absent"))
+
+
+class TestRecordedFields(_TempRegistryCase):
+    """AC: records provider, model, project, role, namespace/profile, workflow,
+    schedule, timeout, enabled state, and scheduler metadata."""
+
+    def test_every_required_field_is_recorded(self):
+        registry = self.parse(_document())
+        worker = registry.workers[0]
+        self.assertEqual(worker.provider, "claude")
+        self.assertEqual(worker.model, "claude-opus-4-8")
+        self.assertEqual(worker.project, "gitea-tools")
+        self.assertEqual(worker.role, "author")
+        self.assertEqual(worker.namespace, "gitea-author")
+        self.assertEqual(worker.profile, "prgs-author")
+        self.assertEqual(worker.workflow, "skills/llm-project-workflow/workflows/work-issue.md")
+        self.assertEqual(worker.schedule.kind, "cron")
+        self.assertEqual(worker.schedule.expression, "0 * * * *")
+        self.assertEqual(worker.timeout_seconds, 3600)
+        self.assertTrue(worker.enabled)
+        self.assertEqual(worker.scheduler.kind, "launchd")
+        self.assertEqual(worker.scheduler.label, "cc.prgs.claude.author")
+
+    def test_each_required_field_is_individually_required(self):
+        for field in (
+            "provider", "model", "project", "role", "namespace",
+            "profile", "workflow", "schedule", "timeout_seconds",
+            "enabled", "scheduler", "id", "display_name",
+        ):
+            with self.subTest(field=field):
+                worker = _worker()
+                worker.pop(field)
+                with self.assertRaises(RegistryValidationError):
+                    self.parse(_document(workers=[worker]))
+
+    def test_all_sanctioned_roles_are_accepted(self):
+        for role in ALLOWED_ROLES:
+            with self.subTest(role=role):
+                registry = self.parse(_document(workers=[_worker(role=role)]))
+                self.assertEqual(registry.workers[0].role, role)
+
+    def test_unsanctioned_role_is_refused(self):
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(workers=[_worker(role="admin")]))
+        self.assertIn("role must be one of", str(ctx.exception))
+
+    def test_worker_dict_round_trips_every_field(self):
+        registry = self.parse(_document())
+        encoded = worker_to_dict(registry.workers[0])
+        self.assertEqual(encoded, _worker())
+        json.dumps(encoded)  # must stay JSON-safe for the #799 API
+
+
+class TestSchemaValidation(_TempRegistryCase):
+    """AC: supports schema validation — and fails closed."""
+
+    def test_unsupported_version_is_refused(self):
+        with self.assertRaises(RegistryValidationError):
+            self.parse(_document(version=2))
+
+    def test_root_must_be_an_object(self):
+        with self.assertRaises(RegistryValidationError):
+            validate_payload([], source_path=self.path)
+
+    def test_providers_must_be_non_empty(self):
+        with self.assertRaises(RegistryValidationError):
+            self.parse(_document(providers=[]))
+
+    def test_unknown_top_level_field_is_refused(self):
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(fleet=[]))
+        self.assertIn("unknown fields", str(ctx.exception))
+
+    def test_unknown_worker_field_is_refused_not_ignored(self):
+        # A typo'd field must not be silently dropped: "timeout_second" would
+        # otherwise read as "no timeout declared".
+        worker = _worker()
+        worker["timeout_second"] = 30
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(workers=[worker]))
+        self.assertIn("timeout_second", str(ctx.exception))
+
+    def test_credentials_are_refused_anywhere_in_the_document(self):
+        for label, mutate in (
+            ("provider.api_token", lambda doc: doc["providers"][0].__setitem__("api_token", "x")),
+            ("worker.password", lambda doc: doc["workers"][0].__setitem__("password", "x")),
+            ("root.secret", lambda doc: doc.__setitem__("secret", "x")),
+        ):
+            with self.subTest(field=label):
+                document = _document()
+                mutate(document)
+                with self.assertRaises(ValueError) as ctx:
+                    self.parse(document)
+                self.assertIn("credential", str(ctx.exception).lower())
+
+    def test_duplicate_worker_id_is_refused(self):
+        workers = [_worker("dup"), _worker("dup", scheduler={"kind": "manual"})]
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(workers=workers))
+        self.assertIn("duplicate worker id", str(ctx.exception))
+
+    def test_duplicate_provider_id_is_refused(self):
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(providers=[_provider("claude"), _provider("claude")], workers=[]))
+        self.assertIn("duplicate provider id", str(ctx.exception))
+
+    def test_duplicate_launchagent_label_is_refused(self):
+        # Two workers sharing a label would silently overwrite each other's agent.
+        workers = [
+            _worker("a", scheduler={"kind": "launchd", "label": "cc.prgs.same"}),
+            _worker("b", scheduler={"kind": "launchd", "label": "cc.prgs.same"}),
+        ]
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(workers=workers))
+        self.assertIn("duplicate scheduler label", str(ctx.exception))
+
+    def test_manual_scheduler_needs_no_label_and_many_may_coexist(self):
+        workers = [
+            _worker("a", scheduler={"kind": "manual"}),
+            _worker("b", scheduler={"kind": "manual"}),
+        ]
+        registry = self.parse(_document(workers=workers))
+        self.assertEqual([w.scheduler.label for w in registry.workers], [None, None])
+
+    def test_launchd_scheduler_requires_a_label(self):
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(workers=[_worker(scheduler={"kind": "launchd"})]))
+        self.assertIn("label is required", str(ctx.exception))
+
+    def test_unknown_scheduler_kind_is_refused(self):
+        with self.assertRaises(RegistryValidationError):
+            self.parse(_document(workers=[_worker(scheduler={"kind": "systemd", "label": "x"})]))
+
+    def test_timeout_must_be_a_positive_bounded_integer(self):
+        for bad in (0, -1, "3600", 1.5, True, 86_401):
+            with self.subTest(timeout=bad):
+                with self.assertRaises(RegistryValidationError):
+                    self.parse(_document(workers=[_worker(timeout_seconds=bad)]))
+
+    def test_enabled_must_be_a_real_boolean(self):
+        for bad in ("true", 1, None):
+            with self.subTest(enabled=bad):
+                with self.assertRaises(RegistryValidationError):
+                    self.parse(_document(workers=[_worker(enabled=bad)]))
+
+    def test_identifier_shape_is_enforced(self):
+        for bad in ("Claude Author", "-leading", "UPPER", ""):
+            with self.subTest(worker_id=bad):
+                with self.assertRaises(RegistryValidationError):
+                    self.parse(_document(workers=[_worker(bad)]))
+
+
+class TestScheduleValidation(_TempRegistryCase):
+    """Schedules are declarations; next-run computation belongs to #803."""
+
+    def test_interval_schedule_requires_positive_seconds(self):
+        registry = self.parse(
+            _document(workers=[_worker(schedule={"kind": "interval", "seconds": 900})])
+        )
+        self.assertEqual(registry.workers[0].schedule.seconds, 900)
+        with self.assertRaises(RegistryValidationError):
+            self.parse(_document(workers=[_worker(schedule={"kind": "interval"})]))
+        with self.assertRaises(RegistryValidationError):
+            self.parse(_document(workers=[_worker(schedule={"kind": "interval", "seconds": 0})]))
+
+    def test_cron_schedule_requires_five_fields(self):
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(workers=[_worker(schedule={"kind": "cron", "expression": "0 *"})]))
+        self.assertIn("five crontab fields", str(ctx.exception))
+
+    def test_manual_schedule_needs_no_timing(self):
+        registry = self.parse(_document(workers=[_worker(schedule={"kind": "manual"})]))
+        schedule = registry.workers[0].schedule
+        self.assertEqual(schedule.kind, "manual")
+        self.assertIsNone(schedule.seconds)
+        self.assertIsNone(schedule.expression)
+
+    def test_fields_from_the_wrong_kind_are_refused(self):
+        with self.assertRaises(RegistryValidationError) as ctx:
+            self.parse(_document(workers=[_worker(schedule={"kind": "manual", "seconds": 60})]))
+        self.assertIn("not valid for kind", str(ctx.exception))
+
+    def test_unknown_schedule_kind_is_refused(self):
+        with self.assertRaises(RegistryValidationError):
+            self.parse(_document(workers=[_worker(schedule={"kind": "hourly"})]))
+
+
+class TestAtomicPersistence(_TempRegistryCase):
+    """AC: atomic persistence."""
+
+    def test_save_then_load_round_trips(self):
+        registry = self.parse(_document())
+        save_registry(registry, self.path)
+        reloaded = load_registry(self.path)
+        self.assertEqual(
+            [worker_to_dict(w) for w in reloaded.workers],
+            [worker_to_dict(w) for w in registry.workers],
+        )
+
+    def test_save_leaves_no_temp_files_behind(self):
+        registry = self.parse(_document())
+        save_registry(registry, self.path)
+        save_registry(registry, self.path)
+        leftovers = [p.name for p in self.path.parent.iterdir() if p.name.startswith(".")]
+        self.assertEqual(leftovers, [])
+
+    def test_save_refuses_to_persist_an_invalid_document(self):
+        registry = self.parse(_document())
+        broken = WorkerRegistry(
+            version=registry.version,
+            revision=registry.revision,
+            updated_at=registry.updated_at,
+            providers=registry.providers,
+            # A worker whose provider is not declared in the registry.
+            workers=tuple(
+                type(worker)(**{**worker.__dict__, "provider": "vanished"})
+                for worker in registry.workers
+            ),
+            source_path=self.path,
+        )
+        with self.assertRaises(RegistryValidationError):
+            save_registry(broken, self.path)
+        self.assertFalse(self.path.exists(), "invalid save must not create the file")
+
+    def test_document_shape_excludes_local_paths_but_api_shape_includes_it(self):
+        registry = self.parse(_document())
+        self.assertNotIn("source_path", registry_to_document(registry))
+        self.assertEqual(registry_to_dict(registry)["source_path"], str(self.path))
+
+
+class TestVersioningAndRollback(_TempRegistryCase):
+    """AC: versioning and rollback."""
+
+    def _seed(self) -> WorkerRegistry:
+        self.write(_document())
+        return load_registry(self.path)
+
+    def test_revision_increments_on_each_save(self):
+        registry = self._seed()
+        self.assertEqual(registry.revision, 1)
+        second = save_registry(registry, self.path)
+        self.assertEqual(second.revision, 2)
+        third = save_registry(second, self.path)
+        self.assertEqual(third.revision, 3)
+
+    def test_updated_at_is_refreshed_and_utc(self):
+        registry = self._seed()
+        saved = save_registry(registry, self.path)
+        self.assertRegex(saved.updated_at, r"^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}Z$")
+
+    def test_superseded_revisions_are_retained(self):
+        registry = self._seed()
+        second = save_registry(registry, self.path)
+        save_registry(second, self.path)
+        self.assertEqual(list_revisions(self.path), (1, 2))
+        self.assertTrue(history_dir(self.path).is_dir())
+
+    def test_rollback_restores_prior_content_as_a_new_revision(self):
+        self.write(_document(workers=[_worker("original")]))
+        registry = load_registry(self.path)
+
+        changed = WorkerRegistry(
+            version=registry.version,
+            revision=registry.revision,
+            updated_at=registry.updated_at,
+            providers=registry.providers,
+            workers=(),  # operator deletes every worker
+            source_path=self.path,
+        )
+        save_registry(changed, self.path)
+        self.assertEqual(load_registry(self.path).workers, ())
+
+        restored = rollback_to_revision(1, self.path)
+        self.assertEqual([w.id for w in restored.workers], ["original"])
+        # Append-only: the rollback publishes a new head rather than rewinding.
+        self.assertGreater(restored.revision, 2)
+        self.assertEqual([w.id for w in load_registry(self.path).workers], ["original"])
+
+    def test_rollback_to_unknown_revision_fails_closed(self):
+        self._seed()
+        with self.assertRaises(RegistryValidationError) as ctx:
+            rollback_to_revision(99, self.path)
+        self.assertIn("not retained", str(ctx.exception))
+
+    def test_revision_must_be_a_positive_integer(self):
+        for bad in (0, -1, "1", None):
+            with self.subTest(revision=bad):
+                with self.assertRaises(RegistryValidationError):
+                    self.parse(_document(revision=bad))
+
+    def test_history_is_empty_before_any_save(self):
+        self.write(_document())
+        self.assertEqual(list_revisions(self.path), ())
+
+
+if __name__ == "__main__":
+    unittest.main()

webui/data/workers.registry.json

@@ -0,0 +1,57 @@
+{
+  "version": 1,
+  "revision": 1,
+  "updated_at": "2026-07-22T00:00:00Z",
+  "providers": [
+    {
+      "id": "claude",
+      "display_name": "Claude",
+      "vendor": "Anthropic",
+      "executable": "claude",
+      "available": true,
+      "models": [
+        "claude-opus-4-8",
+        "claude-sonnet-5",
+        "claude-haiku-4-5-20251001"
+      ],
+      "notes": "Model list is a declaration. Live enumeration and version inspection belong to the provider adapter framework (#800)."
+    },
+    {
+      "id": "grok",
+      "display_name": "Grok",
+      "vendor": "xAI",
+      "executable": "grok",
+      "available": true,
+      "models": [],
+      "notes": "Models enumerated by the provider adapter (#800); not declared here."
+    },
+    {
+      "id": "codex",
+      "display_name": "Codex",
+      "vendor": "OpenAI",
+      "executable": "codex",
+      "available": true,
+      "models": [],
+      "notes": "Models enumerated by the provider adapter (#800); not declared here."
+    },
+    {
+      "id": "agy",
+      "display_name": "AGY",
+      "vendor": "Antigravity",
+      "executable": "agy",
+      "available": true,
+      "models": [],
+      "notes": "MCP allowlist gating applies to this provider; confirm server-side allowlist before configuring a worker."
+    },
+    {
+      "id": "kimi-k",
+      "display_name": "Kimi K",
+      "vendor": "Moonshot AI",
+      "executable": "kimi",
+      "available": true,
+      "models": [],
+      "notes": "Provider id is kimi-k; the executable on PATH is kimi. Models enumerated by the provider adapter (#800)."
+    }
+  ],
+  "workers": []
+}

webui/project_registry.py

@@ -8,27 +8,7 @@ from dataclasses import dataclass
 from pathlib import Path
 from typing import Any
 
-_FORBIDDEN_EXACT_KEYS = frozenset({
-    "token",
-    "password",
-    "secret",
-    "credential",
-    "auth",
-    "api_key",
-    "api-key",
-})
-_FORBIDDEN_KEY_PREFIXES = ("auth_", "api_key_", "api-key_")
-_FORBIDDEN_KEY_SUFFIXES = ("_token", "_secret", "_password", "_credential", "_auth")
-
-
-def _is_forbidden_key(key: str) -> bool:
-    lowered = key.lower()
-    if lowered in _FORBIDDEN_EXACT_KEYS:
-        return True
-    return (
-        lowered.startswith(_FORBIDDEN_KEY_PREFIXES)
-        or lowered.endswith(_FORBIDDEN_KEY_SUFFIXES)
-    )
+from webui.registry_safety import reject_credential_keys as _reject_credential_keys
 
 _REQUIRED_PROJECT_FIELDS = (
     "id",
@@ -79,18 +59,6 @@ def default_registry_path() -> Path:
     return (Path(__file__).resolve().parent / "data" / "projects.registry.json").resolve()
 
 
-def _reject_credential_keys(obj: Any, *, path: str = "") -> None:
-    if isinstance(obj, dict):
-        for key, value in obj.items():
-            key_path = f"{path}.{key}" if path else key
-            if _is_forbidden_key(key):
-                raise ValueError(f"registry must not store credentials ({key_path})")
-            _reject_credential_keys(value, path=key_path)
-    elif isinstance(obj, list):
-        for index, item in enumerate(obj):
-            _reject_credential_keys(item, path=f"{path}[{index}]")
-
-
 def _parse_onboarding(raw: list[dict[str, Any]] | None) -> tuple[OnboardingStep, ...]:
     if not raw:
         return ()

webui/registry_safety.py

@@ -0,0 +1,52 @@
+"""Shared credential-rejection guard for web UI registries (#427, #798).
+
+Registries are operator-editable declarative files that the web UI loads and,
+for the worker registry, writes back. None of them may ever carry a secret:
+credentials belong in the keychain and reach worker processes through
+environment injection, never through a file the browser layer can read.
+
+The check is structural rather than value-based on purpose. A value scanner has
+to guess what a secret looks like; a key scanner refuses the *shape* of a
+credential field, so an operator cannot introduce one by accident and a later
+loader cannot silently pass one through.
+"""
+
+from __future__ import annotations
+
+from typing import Any
+
+_FORBIDDEN_EXACT_KEYS = frozenset({
+    "token",
+    "password",
+    "secret",
+    "credential",
+    "auth",
+    "api_key",
+    "api-key",
+})
+_FORBIDDEN_KEY_PREFIXES = ("auth_", "api_key_", "api-key_")
+_FORBIDDEN_KEY_SUFFIXES = ("_token", "_secret", "_password", "_credential", "_auth")
+
+
+def is_forbidden_key(key: str) -> bool:
+    """Return True when *key* names a credential field."""
+    lowered = key.lower()
+    if lowered in _FORBIDDEN_EXACT_KEYS:
+        return True
+    return (
+        lowered.startswith(_FORBIDDEN_KEY_PREFIXES)
+        or lowered.endswith(_FORBIDDEN_KEY_SUFFIXES)
+    )
+
+
+def reject_credential_keys(obj: Any, *, path: str = "", subject: str = "registry") -> None:
+    """Raise ValueError when *obj* carries a credential-shaped key at any depth."""
+    if isinstance(obj, dict):
+        for key, value in obj.items():
+            key_path = f"{path}.{key}" if path else key
+            if is_forbidden_key(key):
+                raise ValueError(f"{subject} must not store credentials ({key_path})")
+            reject_credential_keys(value, path=key_path, subject=subject)
+    elif isinstance(obj, list):
+        for index, item in enumerate(obj):
+            reject_credential_keys(item, path=f"{path}[{index}]", subject=subject)

webui/worker_registry.py

@@ -0,0 +1,647 @@
+"""Declarative worker registry and configuration schema (#798, epic #797).
+
+The registry is the single source of truth for the scheduled multi-LLM worker
+fleet. It is a versioned JSON document holding two *separate* entity kinds:
+
+* **Providers** — the LLM runtimes a worker can be built on (Claude, Grok,
+  Codex, AGY, Kimi K). A provider describes the runtime itself: vendor,
+  executable name, models it can serve, and whether it is available on this
+  machine. Providers exist whether or not any worker uses them.
+* **Workers** — a configured *instance*: one provider, one model, one project,
+  one role, one MCP namespace/profile, one workflow, one schedule. Several
+  workers may share a provider; a worker naming an undeclared provider is
+  refused.
+
+Keeping them separate is what lets #799 list all five providers even when a
+provider currently has no configured worker, and it stops provider facts from
+being copied into (and drifting across) every worker record.
+
+Scope boundary. This module owns the data model, its validation, and its
+persistence. It does **not** schedule anything, launch anything, probe provider
+executables, or serve HTTP. Loading a registry never touches a process; the
+live fields a dashboard wants (PID, elapsed time, next run) are derived
+elsewhere (#799, #801, #803, #804) from these declarations.
+
+Safety invariants:
+
+* No credential may be stored (:mod:`webui.registry_safety`), so the registry
+  stays safe to render and to hand to a browser layer.
+* Validation fails closed. Unknown fields are refused rather than ignored, so a
+  typo cannot silently disable a timeout or a role binding.
+* Writes are atomic and every superseded document is retained as a numbered
+  revision, so a bad edit is recoverable by rollback rather than hand-repair.
+"""
+
+from __future__ import annotations
+
+import json
+import os
+import re
+import tempfile
+from dataclasses import dataclass
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any
+
+from webui.registry_safety import reject_credential_keys
+
+SCHEMA_VERSION = 1
+
+#: Roles a worker may hold. These mirror the sanctioned MCP role kinds; a
+#: worker may not invent one, because the role selects the namespace/profile
+#: whose capability gates constrain it.
+ALLOWED_ROLES = ("author", "reviewer", "merger", "reconciler", "cleanup")
+
+#: Scheduler backends the registry can describe. ``manual`` means the worker is
+#: only ever started on request and has no recurring trigger.
+ALLOWED_SCHEDULER_KINDS = ("launchd", "manual")
+
+#: Schedule kinds. Next-run computation belongs to #803; this module only
+#: guarantees the declaration is well formed.
+ALLOWED_SCHEDULE_KINDS = ("interval", "cron", "manual")
+
+_REQUIRED_PROVIDER_FIELDS = ("id", "display_name", "vendor", "executable", "available")
+_OPTIONAL_PROVIDER_FIELDS = ("models", "notes")
+
+_REQUIRED_WORKER_FIELDS = (
+    "id",
+    "display_name",
+    "provider",
+    "model",
+    "project",
+    "role",
+    "namespace",
+    "profile",
+    "workflow",
+    "schedule",
+    "timeout_seconds",
+    "enabled",
+    "scheduler",
+)
+_OPTIONAL_WORKER_FIELDS = ("notes",)
+
+_ID_RE = re.compile(r"^[a-z0-9][a-z0-9._-]*$")
+
+#: Guards against an operator writing a timeout that would let a worker hold a
+#: lease effectively forever. 24h is far above any sanctioned cycle.
+_MAX_TIMEOUT_SECONDS = 86_400
+
+#: How many superseded revisions to retain beside the live file.
+_HISTORY_LIMIT = 20
+
+_TOP_LEVEL_FIELDS = frozenset({"version", "revision", "updated_at", "providers", "workers"})
+
+
+class RegistryValidationError(ValueError):
+    """Raised when a registry document violates the schema."""
+
+
+@dataclass(frozen=True)
+class ProviderRecord:
+    id: str
+    display_name: str
+    vendor: str
+    executable: str
+    available: bool
+    models: tuple[str, ...]
+    notes: str
+
+
+@dataclass(frozen=True)
+class ScheduleSpec:
+    kind: str
+    #: Set for ``interval`` schedules.
+    seconds: int | None
+    #: Set for ``cron`` schedules — a five-field crontab expression.
+    expression: str | None
+
+
+@dataclass(frozen=True)
+class SchedulerSpec:
+    kind: str
+    #: LaunchAgent label; required for ``launchd``, absent for ``manual``.
+    label: str | None
+
+
+@dataclass(frozen=True)
+class WorkerRecord:
+    id: str
+    display_name: str
+    provider: str
+    model: str
+    project: str
+    role: str
+    namespace: str
+    profile: str
+    workflow: str
+    schedule: ScheduleSpec
+    timeout_seconds: int
+    enabled: bool
+    scheduler: SchedulerSpec
+    notes: str
+
+
+@dataclass(frozen=True)
+class WorkerRegistry:
+    version: int
+    revision: int
+    updated_at: str
+    providers: tuple[ProviderRecord, ...]
+    workers: tuple[WorkerRecord, ...]
+    source_path: Path
+
+
+# ── paths ────────────────────────────────────────────────────────────────────
+
+
+def default_registry_path() -> Path:
+    """Location of the packaged worker registry, overridable for tests/deploys."""
+    override = os.environ.get("WEBUI_WORKER_REGISTRY", "").strip()
+    if override:
+        return Path(override).expanduser().resolve()
+    return (Path(__file__).resolve().parent / "data" / "workers.registry.json").resolve()
+
+
+def history_dir(path: Path | None = None) -> Path:
+    """Directory holding superseded revisions of *path*."""
+    source = (path or default_registry_path()).resolve()
+    return source.parent / f"{source.name}.history"
+
+
+# ── field helpers ────────────────────────────────────────────────────────────
+
+
+def _require_exact_fields(
+    raw: Any,
+    *,
+    required: tuple[str, ...],
+    optional: tuple[str, ...],
+    subject: str,
+) -> dict[str, Any]:
+    if not isinstance(raw, dict):
+        raise RegistryValidationError(f"{subject} must be an object")
+    missing = [field for field in required if field not in raw]
+    if missing:
+        raise RegistryValidationError(
+            f"{subject} missing required fields: {', '.join(sorted(missing))}"
+        )
+    unknown = sorted(set(raw) - set(required) - set(optional))
+    if unknown:
+        # Fail closed: silently dropping an unrecognized key is how a typo'd
+        # "timeout_second" ends up meaning "no timeout".
+        raise RegistryValidationError(f"{subject} has unknown fields: {', '.join(unknown)}")
+    return raw
+
+
+def _require_identifier(value: Any, *, subject: str) -> str:
+    text = str(value).strip()
+    if not _ID_RE.match(text):
+        raise RegistryValidationError(
+            f"{subject} must be lowercase alphanumeric with '.', '_', or '-' (got {value!r})"
+        )
+    return text
+
+
+def _require_text(value: Any, *, subject: str) -> str:
+    if not isinstance(value, str):
+        raise RegistryValidationError(f"{subject} must be a string (got {value!r})")
+    text = value.strip()
+    if not text:
+        raise RegistryValidationError(f"{subject} must be a non-empty string")
+    return text
+
+
+def _require_bool(value: Any, *, subject: str) -> bool:
+    if not isinstance(value, bool):
+        raise RegistryValidationError(f"{subject} must be a boolean (got {value!r})")
+    return value
+
+
+def _require_positive_int(value: Any, *, subject: str, maximum: int | None = None) -> int:
+    if isinstance(value, bool) or not isinstance(value, int):
+        raise RegistryValidationError(f"{subject} must be an integer (got {value!r})")
+    if value <= 0:
+        raise RegistryValidationError(f"{subject} must be greater than zero (got {value})")
+    if maximum is not None and value > maximum:
+        raise RegistryValidationError(f"{subject} must not exceed {maximum} (got {value})")
+    return value
+
+
+# ── parsing ──────────────────────────────────────────────────────────────────
+
+
+def _parse_provider(raw: Any) -> ProviderRecord:
+    data = _require_exact_fields(
+        raw,
+        required=_REQUIRED_PROVIDER_FIELDS,
+        optional=_OPTIONAL_PROVIDER_FIELDS,
+        subject="provider",
+    )
+    provider_id = _require_identifier(data["id"], subject="provider.id")
+
+    models_raw = data.get("models") or []
+    if not isinstance(models_raw, list):
+        raise RegistryValidationError(f"provider[{provider_id}].models must be an array")
+    models = tuple(
+        _require_text(item, subject=f"provider[{provider_id}].models[]") for item in models_raw
+    )
+
+    return ProviderRecord(
+        id=provider_id,
+        display_name=_require_text(
+            data["display_name"], subject=f"provider[{provider_id}].display_name"
+        ),
+        vendor=_require_text(data["vendor"], subject=f"provider[{provider_id}].vendor"),
+        executable=_require_text(data["executable"], subject=f"provider[{provider_id}].executable"),
+        available=_require_bool(data["available"], subject=f"provider[{provider_id}].available"),
+        models=models,
+        notes=str(data.get("notes") or "").strip(),
+    )
+
+
+def _parse_schedule(raw: Any, *, subject: str) -> ScheduleSpec:
+    if not isinstance(raw, dict):
+        raise RegistryValidationError(f"{subject} must be an object")
+    kind = _require_text(raw.get("kind"), subject=f"{subject}.kind")
+    if kind not in ALLOWED_SCHEDULE_KINDS:
+        raise RegistryValidationError(
+            f"{subject}.kind must be one of {', '.join(ALLOWED_SCHEDULE_KINDS)} (got {kind!r})"
+        )
+
+    seconds: int | None = None
+    expression: str | None = None
+
+    if kind == "interval":
+        if "seconds" not in raw:
+            raise RegistryValidationError(f"{subject}.seconds is required for interval schedules")
+        seconds = _require_positive_int(raw["seconds"], subject=f"{subject}.seconds")
+    elif kind == "cron":
+        if "expression" not in raw:
+            raise RegistryValidationError(f"{subject}.expression is required for cron schedules")
+        expression = _require_text(raw["expression"], subject=f"{subject}.expression")
+        if len(expression.split()) != 5:
+            raise RegistryValidationError(
+                f"{subject}.expression must have five crontab fields (got {expression!r})"
+            )
+
+    allowed = {"kind"}
+    if kind == "interval":
+        allowed.add("seconds")
+    elif kind == "cron":
+        allowed.add("expression")
+    unknown = sorted(set(raw) - allowed)
+    if unknown:
+        raise RegistryValidationError(
+            f"{subject} has fields not valid for kind {kind!r}: {', '.join(unknown)}"
+        )
+
+    return ScheduleSpec(kind=kind, seconds=seconds, expression=expression)
+
+
+def _parse_scheduler(raw: Any, *, subject: str) -> SchedulerSpec:
+    if not isinstance(raw, dict):
+        raise RegistryValidationError(f"{subject} must be an object")
+    kind = _require_text(raw.get("kind"), subject=f"{subject}.kind")
+    if kind not in ALLOWED_SCHEDULER_KINDS:
+        raise RegistryValidationError(
+            f"{subject}.kind must be one of {', '.join(ALLOWED_SCHEDULER_KINDS)} (got {kind!r})"
+        )
+
+    label: str | None = None
+    if kind == "launchd":
+        if "label" not in raw:
+            raise RegistryValidationError(f"{subject}.label is required for launchd schedulers")
+        label = _require_text(raw["label"], subject=f"{subject}.label")
+
+    allowed = {"kind"}
+    if kind == "launchd":
+        allowed.add("label")
+    unknown = sorted(set(raw) - allowed)
+    if unknown:
+        raise RegistryValidationError(
+            f"{subject} has fields not valid for kind {kind!r}: {', '.join(unknown)}"
+        )
+
+    return SchedulerSpec(kind=kind, label=label)
+
+
+def _parse_worker(raw: Any) -> WorkerRecord:
+    data = _require_exact_fields(
+        raw,
+        required=_REQUIRED_WORKER_FIELDS,
+        optional=_OPTIONAL_WORKER_FIELDS,
+        subject="worker",
+    )
+    worker_id = _require_identifier(data["id"], subject="worker.id")
+
+    role = _require_text(data["role"], subject=f"worker[{worker_id}].role")
+    if role not in ALLOWED_ROLES:
+        raise RegistryValidationError(
+            f"worker[{worker_id}].role must be one of {', '.join(ALLOWED_ROLES)} (got {role!r})"
+        )
+
+    return WorkerRecord(
+        id=worker_id,
+        display_name=_require_text(
+            data["display_name"], subject=f"worker[{worker_id}].display_name"
+        ),
+        provider=_require_identifier(data["provider"], subject=f"worker[{worker_id}].provider"),
+        model=_require_text(data["model"], subject=f"worker[{worker_id}].model"),
+        project=_require_text(data["project"], subject=f"worker[{worker_id}].project"),
+        role=role,
+        namespace=_require_text(data["namespace"], subject=f"worker[{worker_id}].namespace"),
+        profile=_require_text(data["profile"], subject=f"worker[{worker_id}].profile"),
+        workflow=_require_text(data["workflow"], subject=f"worker[{worker_id}].workflow"),
+        schedule=_parse_schedule(data["schedule"], subject=f"worker[{worker_id}].schedule"),
+        timeout_seconds=_require_positive_int(
+            data["timeout_seconds"],
+            subject=f"worker[{worker_id}].timeout_seconds",
+            maximum=_MAX_TIMEOUT_SECONDS,
+        ),
+        enabled=_require_bool(data["enabled"], subject=f"worker[{worker_id}].enabled"),
+        scheduler=_parse_scheduler(data["scheduler"], subject=f"worker[{worker_id}].scheduler"),
+        notes=str(data.get("notes") or "").strip(),
+    )
+
+
+def _require_unique(values: list[str], *, subject: str) -> None:
+    seen: set[str] = set()
+    for value in values:
+        if value in seen:
+            raise RegistryValidationError(f"duplicate {subject}: {value}")
+        seen.add(value)
+
+
+def validate_payload(payload: Any, *, source_path: Path) -> WorkerRegistry:
+    """Validate a decoded registry document and return the typed registry.
+
+    Raises :class:`RegistryValidationError` on any violation; never partially
+    accepts a document.
+    """
+    if not isinstance(payload, dict):
+        raise RegistryValidationError("registry root must be an object")
+
+    version = payload.get("version")
+    if version != SCHEMA_VERSION:
+        raise RegistryValidationError(f"unsupported registry version: {version!r}")
+
+    reject_credential_keys(payload, subject="worker registry")
+
+    unknown = sorted(set(payload) - _TOP_LEVEL_FIELDS)
+    if unknown:
+        raise RegistryValidationError(f"registry has unknown fields: {', '.join(unknown)}")
+
+    revision = _require_positive_int(payload.get("revision"), subject="revision")
+    updated_at = _require_text(payload.get("updated_at"), subject="updated_at")
+
+    providers_raw = payload.get("providers")
+    if not isinstance(providers_raw, list) or not providers_raw:
+        raise RegistryValidationError("providers must be a non-empty array")
+    providers = tuple(_parse_provider(item) for item in providers_raw)
+    _require_unique([provider.id for provider in providers], subject="provider id")
+
+    workers_raw = payload.get("workers")
+    if not isinstance(workers_raw, list):
+        raise RegistryValidationError("workers must be an array")
+    workers = tuple(_parse_worker(item) for item in workers_raw)
+    _require_unique([worker.id for worker in workers], subject="worker id")
+
+    # Referential integrity: a worker naming an undeclared provider would look
+    # configured while being unrunnable, which is exactly the ambiguous
+    # ownership the epic requires to fail closed.
+    known_providers = {provider.id for provider in providers}
+    for worker in workers:
+        if worker.provider not in known_providers:
+            raise RegistryValidationError(
+                f"worker[{worker.id}].provider references unknown provider {worker.provider!r}"
+            )
+
+    # A LaunchAgent label identifies a job to launchd; two workers sharing one
+    # would silently overwrite each other's agent.
+    _require_unique(
+        [worker.scheduler.label for worker in workers if worker.scheduler.label],
+        subject="scheduler label",
+    )
+
+    return WorkerRegistry(
+        version=version,
+        revision=revision,
+        updated_at=updated_at,
+        providers=providers,
+        workers=workers,
+        source_path=source_path,
+    )
+
+
+def load_registry(path: Path | None = None) -> WorkerRegistry:
+    """Load and validate the worker registry from disk."""
+    source = (path or default_registry_path()).resolve()
+    payload = json.loads(source.read_text(encoding="utf-8"))
+    return validate_payload(payload, source_path=source)
+
+
+# ── serialization ────────────────────────────────────────────────────────────
+
+
+def provider_to_dict(provider: ProviderRecord) -> dict[str, Any]:
+    return {
+        "id": provider.id,
+        "display_name": provider.display_name,
+        "vendor": provider.vendor,
+        "executable": provider.executable,
+        "available": provider.available,
+        "models": list(provider.models),
+        "notes": provider.notes,
+    }
+
+
+def _schedule_to_dict(schedule: ScheduleSpec) -> dict[str, Any]:
+    payload: dict[str, Any] = {"kind": schedule.kind}
+    if schedule.kind == "interval":
+        payload["seconds"] = schedule.seconds
+    elif schedule.kind == "cron":
+        payload["expression"] = schedule.expression
+    return payload
+
+
+def _scheduler_to_dict(scheduler: SchedulerSpec) -> dict[str, Any]:
+    payload: dict[str, Any] = {"kind": scheduler.kind}
+    if scheduler.kind == "launchd":
+        payload["label"] = scheduler.label
+    return payload
+
+
+def worker_to_dict(worker: WorkerRecord) -> dict[str, Any]:
+    return {
+        "id": worker.id,
+        "display_name": worker.display_name,
+        "provider": worker.provider,
+        "model": worker.model,
+        "project": worker.project,
+        "role": worker.role,
+        "namespace": worker.namespace,
+        "profile": worker.profile,
+        "workflow": worker.workflow,
+        "schedule": _schedule_to_dict(worker.schedule),
+        "timeout_seconds": worker.timeout_seconds,
+        "enabled": worker.enabled,
+        "scheduler": _scheduler_to_dict(worker.scheduler),
+        "notes": worker.notes,
+    }
+
+
+def registry_to_document(registry: WorkerRegistry) -> dict[str, Any]:
+    """Serialize to the on-disk document shape (no local paths embedded)."""
+    return {
+        "version": registry.version,
+        "revision": registry.revision,
+        "updated_at": registry.updated_at,
+        "providers": [provider_to_dict(provider) for provider in registry.providers],
+        "workers": [worker_to_dict(worker) for worker in registry.workers],
+    }
+
+
+def registry_to_dict(registry: WorkerRegistry) -> dict[str, Any]:
+    """Serialize for JSON API responses (adds the resolved source path)."""
+    document = registry_to_document(registry)
+    document["source_path"] = str(registry.source_path)
+    return document
+
+
+def find_worker(registry: WorkerRegistry, worker_id: str) -> WorkerRecord | None:
+    for worker in registry.workers:
+        if worker.id == worker_id:
+            return worker
+    return None
+
+
+def find_provider(registry: WorkerRegistry, provider_id: str) -> ProviderRecord | None:
+    for provider in registry.providers:
+        if provider.id == provider_id:
+            return provider
+    return None
+
+
+def workers_for_provider(registry: WorkerRegistry, provider_id: str) -> tuple[WorkerRecord, ...]:
+    return tuple(worker for worker in registry.workers if worker.provider == provider_id)
+
+
+# ── persistence ──────────────────────────────────────────────────────────────
+
+
+def _utc_now() -> str:
+    return datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
+
+
+def _atomic_write(path: Path, payload: str) -> None:
+    """Write *payload* to *path* atomically: temp file in the same dir, fsync, replace."""
+    parent = path.parent
+    parent.mkdir(parents=True, exist_ok=True)
+    fd, temp_path = tempfile.mkstemp(prefix=f".{path.name}-", suffix=".tmp", 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 _revision_path(directory: Path, revision: int) -> Path:
+    return directory / f"rev-{revision:06d}.json"
+
+
+def _prune_history(path: Path) -> None:
+    directory = history_dir(path)
+    revisions = list_revisions(path)
+    excess = len(revisions) - _HISTORY_LIMIT
+    for revision in revisions[: max(0, excess)]:
+        _revision_path(directory, revision).unlink(missing_ok=True)
+
+
+def _archive_current(path: Path) -> int | None:
+    """Copy the live document into the history dir under its own revision number."""
+    if not path.exists():
+        return None
+    try:
+        existing = json.loads(path.read_text(encoding="utf-8"))
+        revision = int(existing.get("revision", 0))
+    except (json.JSONDecodeError, TypeError, ValueError, AttributeError):
+        # An unreadable live file has no trustworthy revision number to file it
+        # under, so it cannot join the history chain.
+        return None
+    if revision <= 0:
+        return None
+    _atomic_write(
+        _revision_path(history_dir(path), revision),
+        json.dumps(existing, indent=2, sort_keys=True) + "\n",
+    )
+    _prune_history(path)
+    return revision
+
+
+def list_revisions(path: Path | None = None) -> tuple[int, ...]:
+    """Revision numbers retained in history for *path*, oldest first."""
+    directory = history_dir(path)
+    if not directory.is_dir():
+        return ()
+    revisions: list[int] = []
+    for entry in directory.glob("rev-*.json"):
+        try:
+            revisions.append(int(entry.stem.split("-", 1)[1]))
+        except (IndexError, ValueError):
+            continue
+    return tuple(sorted(revisions))
+
+
+def save_registry(
+    registry: WorkerRegistry,
+    path: Path | None = None,
+    *,
+    updated_at: str | None = None,
+) -> WorkerRegistry:
+    """Validate, archive the superseded revision, then atomically persist a new one.
+
+    The stored revision is always the previous revision plus one, so a reader
+    can tell two documents apart even when their content is otherwise equal.
+    Returns the registry exactly as persisted.
+    """
+    target = (path or registry.source_path or default_registry_path()).resolve()
+
+    document = registry_to_document(registry)
+    # Re-validate before writing: a registry assembled in memory has not
+    # necessarily been through the loader.
+    validate_payload(document, source_path=target)
+
+    archived = _archive_current(target)
+    document["revision"] = (archived + 1) if archived is not None else registry.revision
+    document["updated_at"] = updated_at or _utc_now()
+
+    persisted = validate_payload(document, source_path=target)
+    _atomic_write(target, json.dumps(document, indent=2, sort_keys=True) + "\n")
+    return persisted
+
+
+def rollback_to_revision(revision: int, path: Path | None = None) -> WorkerRegistry:
+    """Restore a retained *revision* as a new head revision.
+
+    History is append-only: rolling back does not delete the revisions in
+    between, it republishes the chosen one under the next revision number, so a
+    rollback is itself reversible.
+    """
+    target = (path or default_registry_path()).resolve()
+    snapshot_path = _revision_path(history_dir(target), revision)
+    if not snapshot_path.exists():
+        available = ", ".join(str(item) for item in list_revisions(target)) or "(none)"
+        raise RegistryValidationError(
+            f"revision {revision} is not retained for {target.name}; available: {available}"
+        )
+
+    payload = json.loads(snapshot_path.read_text(encoding="utf-8"))
+    restored = validate_payload(payload, source_path=target)
+    return save_registry(restored, target)