Merge pull request 'fix(mcp): bind post-merge moot-lease cleanup to the reconciler capability (Closes #745)' (#746) from fix/issue-745-reconciler-moot-lease-gate into master

#753 added the dead-session author-lock recovery assessor, but no sanctioned
recovery could ever reach the lock write. A dead-session lock is by construction
a lock for work that already has an open PR, and the #400 duplicate-work gate
blocked unconditionally on any linked open PR. gitea_lock_issue computed
recovery_sanctioned, then discarded it one gate later:

    open PR #750 already covers issue #749 (fail closed)

Because the recovered lock was never persisted, it stayed non-live, so
_prove_author_ownership_for_pr found no live author lock and
gitea_update_pr_branch_by_merge failed closed too. Both blockers had a single
cause, so only the first one is fixed here.

issue_lock_recovery.owning_pr_recovery_evidence distils a granted assessment
into the minimum evidence the duplicate gate needs: issue, branch, owning PR
number, and head. It returns None unless the outcome is RECOVERY_SANCTIONED and
the assessment's own evidence agrees that local head == remote head == PR head,
so a partial or hand-built assessment cannot authorize anything.

The duplicate gate takes that evidence and re-checks every element against the
live PR list it was handed: exactly one linked open PR, matching number, head
branch, head SHA, and locked branch. Only that exact self-owned PR is exempt.
A different PR, several linked PRs, a different branch or head, or evidence for
another issue all keep failing closed, with a diagnostic naming which element
disagreed. The competing-branch, claim-lease and commit/push/create_pr arms are
untouched, and with no evidence the gate behaves exactly as before.

Ownership proof needed no change: once the recovered lock persists under the
live session pid, _prove_author_ownership_for_pr matches it as before.

Out of scope: the PHASE_COMMIT/PHASE_PUSH/PHASE_CREATE_PR rechecks still block
on a linked open PR for every author, recovered or not. That is pre-existing
#400 behavior, unrelated to lock recovery, and #755 does not cover it.

Tests
- tests/test_issue_755_owning_pr_recovery.py (new, 31 tests) drives the real
  mcp_server.gitea_lock_issue handler, not only the pure assessor: sanctioned
  recovery relocks, persists a live lease with truthful dead_session_recovery
  provenance, and satisfies _prove_author_ownership_for_pr; competing PR,
  multiple linked PRs, mismatched head/branch/worktree, dirty worktree, live
  prior pid, and a fresh claim with no prior lock all stay blocked.
- Neutralizing owning_pr_recovery_evidence regresses all three success tests to
  the exact production error above, so the coverage is load-bearing.
- Focused suites (755, 753, duplicate gates, lock registration, provenance) —
  102 passed.
- Lock/ownership/worktree/handoff suites — 172 passed, 16 subtests.
- Full suite tests/ — 3529 passed, 6 skipped, 2 failed, 365 subtests.
- The same 2 failures reproduce on a pristine detached worktree at
  08f67007c5 (3498 passed, 6 skipped, 2 failed):
  test_issue_702_review_findings_f1_f6 F1 recovery-before-probe and
  test_reconciler_supersession_close org/repo forwarding. Pre-existing.
- git diff --check clean.

Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_017BNsk3KUuFchaZyPJsCxjk

create_issue_bootstrap.py

@@ -0,0 +1,227 @@
+"""Sanctioned pre-issue bootstrap for ``create_issue`` (#749).
+
+``gitea_create_issue`` is a pure remote mutation: it creates a tracking issue
+and writes nothing to the local working tree. The issue-first gate forbids
+creating ``branches/issue-<N>-*`` before the issue number exists, while the
+#274 branches-only guard previously demanded that worktree first — a deadlock.
+
+This module defines a **narrow, phase-scoped** exemption:
+
+* Only tasks in :data:`CREATE_ISSUE_TASKS` may use it.
+* Only the **canonical control checkout** may be used (never an arbitrary
+  directory, unrelated worktree, or foreign clone).
+* The control checkout must be clean, on an accepted base branch, and
+  base-equivalent to live master when a remote tip is known.
+* Every post-creation author mutation keeps the ordinary ``branches/`` rule.
+
+The exemption cannot widen: unknown tasks, dirty roots, drifted HEADs, non-base
+branches, and non-control workspaces fall through to the existing fail-closed
+guards.
+"""
+
+from __future__ import annotations
+
+import os
+from typing import Any
+
+from author_mutation_worktree import BASE_BRANCHES, is_path_under_branches
+from reviewer_worktree import parse_dirty_tracked_files
+
+CREATE_ISSUE_TASKS = frozenset({"create_issue", "gitea_create_issue"})
+
+# Satisfiable before an issue number exists — never names issue-<N>.
+EXACT_NEXT_ACTION_BOOTSTRAP = (
+    "Restore the canonical control checkout to a clean accepted base branch "
+    "(master/main/dev) that matches live master, with no tracked local edits "
+    "and no detached HEAD. Re-resolve the exact create_issue task, then re-run "
+    "gitea_create_issue from that clean control checkout. Do not create "
+    "branches/issue-<N>-* worktrees, dummy directories, or borrow unrelated "
+    "worktrees before the issue exists."
+)
+
+EXACT_NEXT_ACTION_POST_CREATE = (
+    "After the issue exists: create a registered worktree under "
+    "branches/issue-<N>-* from clean master, claim/lock the issue, set "
+    "GITEA_AUTHOR_WORKTREE / worktree_path to that path, then continue author "
+    "mutations from the issue-backed worktree only."
+)
+
+
+def is_create_issue_task(task: str | None) -> bool:
+    """True when *task* is the create_issue mutation (or tool alias)."""
+    return (task or "").strip() in CREATE_ISSUE_TASKS
+
+
+def assess_create_issue_bootstrap(
+    *,
+    workspace_path: str,
+    canonical_repo_root: str,
+    current_branch: str | None = None,
+    head_sha: str | None = None,
+    porcelain_status: str = "",
+    remote_master_sha: str | None = None,
+    task: str | None = None,
+) -> dict[str, Any]:
+    """Assess whether create_issue may proceed from the control checkout.
+
+    Returns a structured assessment:
+
+    * ``not_applicable`` — not a create_issue task, or workspace is already a
+      ``branches/`` worktree (use ordinary guards).
+    * ``allowed`` — create_issue bootstrap may proceed from this control root.
+    * ``block`` — create_issue was attempted from control checkout but gates
+      failed (dirty, wrong branch, base race, etc.).
+    """
+    reasons: list[str] = []
+    root = os.path.realpath(canonical_repo_root or "")
+    workspace = os.path.realpath(workspace_path or root or ".")
+    branch = (current_branch or "").strip()
+    dirty = parse_dirty_tracked_files(porcelain_status or "")
+    under_branches = is_path_under_branches(workspace, root) if root else False
+
+    if not is_create_issue_task(task):
+        return _result(
+            not_applicable=True,
+            allowed=False,
+            block=False,
+            reasons=["task is not create_issue"],
+            workspace=workspace,
+            root=root,
+            branch=branch,
+            dirty=dirty,
+            under_branches=under_branches,
+        )
+
+    # Registered branches/ worktrees keep the normal path (no bootstrap).
+    if under_branches:
+        return _result(
+            not_applicable=True,
+            allowed=False,
+            block=False,
+            reasons=["workspace is under branches/; ordinary #274 path applies"],
+            workspace=workspace,
+            root=root,
+            branch=branch,
+            dirty=dirty,
+            under_branches=True,
+        )
+
+    # Only the exact canonical control checkout is eligible.
+    if not root or workspace != root:
+        reasons.append(
+            "create_issue bootstrap requires the canonical control checkout; "
+            f"workspace '{workspace}' is not the repository root '{root or '(unknown)'}'"
+        )
+        return _result(
+            not_applicable=False,
+            allowed=False,
+            block=True,
+            reasons=reasons,
+            workspace=workspace,
+            root=root,
+            branch=branch,
+            dirty=dirty,
+            under_branches=False,
+            exact_next_action=EXACT_NEXT_ACTION_BOOTSTRAP,
+        )
+
+    if dirty:
+        reasons.append(
+            "create_issue bootstrap blocked: control checkout has tracked local "
+            f"edits (dirty files: {', '.join(dirty)})"
+        )
+
+    if not branch:
+        reasons.append(
+            "create_issue bootstrap blocked: control checkout is detached HEAD; "
+            "expected an accepted base branch (master/main/dev)"
+        )
+    elif branch not in BASE_BRANCHES:
+        reasons.append(
+            f"create_issue bootstrap blocked: control checkout branch '{branch}' "
+            f"is not an accepted base branch ({'/'.join(sorted(BASE_BRANCHES))})"
+        )
+
+    remote_tip = (remote_master_sha or "").strip() or None
+    local_tip = (head_sha or "").strip() or None
+    if remote_tip and local_tip and remote_tip != local_tip:
+        reasons.append(
+            "create_issue bootstrap blocked: control checkout HEAD does not match "
+            f"live master (HEAD {local_tip[:12]}, live master {remote_tip[:12]})"
+        )
+
+    if reasons:
+        return _result(
+            not_applicable=False,
+            allowed=False,
+            block=True,
+            reasons=reasons,
+            workspace=workspace,
+            root=root,
+            branch=branch or None,
+            dirty=dirty,
+            under_branches=False,
+            exact_next_action=EXACT_NEXT_ACTION_BOOTSTRAP,
+        )
+
+    return _result(
+        not_applicable=False,
+        allowed=True,
+        block=False,
+        reasons=[],
+        workspace=workspace,
+        root=root,
+        branch=branch or None,
+        dirty=dirty,
+        under_branches=False,
+        exact_next_action=EXACT_NEXT_ACTION_POST_CREATE,
+        bootstrap_path="clean_canonical_control_checkout",
+    )
+
+
+def format_create_issue_bootstrap_error(assessment: dict[str, Any]) -> str:
+    """RuntimeError / typed-block message for a failed bootstrap assessment."""
+    reasons = "; ".join(
+        assessment.get("reasons") or ["create_issue bootstrap failed"]
+    )
+    next_action = (
+        assessment.get("exact_next_action") or EXACT_NEXT_ACTION_BOOTSTRAP
+    )
+    root = assessment.get("canonical_repo_root") or "(unknown)"
+    workspace = assessment.get("workspace_path") or "(unknown)"
+    return (
+        f"Create-issue bootstrap guard (#749): {reasons}. "
+        f"canonical repository root: {root}; workspace: {workspace}. "
+        f"exact_next_action: {next_action}"
+    )
+
+
+def _result(
+    *,
+    not_applicable: bool,
+    allowed: bool,
+    block: bool,
+    reasons: list[str],
+    workspace: str,
+    root: str,
+    branch: str | None,
+    dirty: list[str],
+    under_branches: bool,
+    exact_next_action: str | None = None,
+    bootstrap_path: str | None = None,
+) -> dict[str, Any]:
+    return {
+        "not_applicable": not_applicable,
+        "allowed": allowed,
+        "block": block,
+        "proven": allowed and not block,
+        "reasons": list(reasons),
+        "workspace_path": workspace,
+        "canonical_repo_root": root,
+        "current_branch": branch,
+        "dirty_files": list(dirty),
+        "under_branches": under_branches,
+        "exact_next_action": exact_next_action,
+        "bootstrap_path": bootstrap_path,
+        "task_scope": "create_issue_only",
+    }

docs/gitea-execution-profiles.md

@@ -317,6 +317,44 @@ Least-privilege constraints:
   canonical names such as `gitea.pr.close` (never bare `pr.close` /
   `issue.close`, which the production normalizer rejects or drops).
 
+### Post-merge moot-lease cleanup ownership (`gitea.pr.comment`)
+
+Neutralising a reviewer lease left behind on an already-merged/closed PR is
+reconciliation work too. `task_capability_map` maps
+`cleanup_post_merge_moot_lease` — and its tool-name alias
+`gitea_cleanup_post_merge_moot_lease` — to role `reconciler` with permission
+`gitea.pr.comment` (#745). Both names carry the **same** contract.
+
+The permission alone is deliberately not sufficient: author, reviewer and
+merger profiles all hold `gitea.pr.comment` for ordinary PR discussion, so the
+role gate — not the permission gate — is what keeps the terminal lease marker
+reconciler-owned.
+
+`gitea_cleanup_post_merge_moot_lease` splits its two modes on purpose:
+
+- **`apply=false` (assessment) requires only `gitea.read`, with no role gate.**
+  This matches `gitea_cleanup_stale_review_decision_lock` and
+  `gitea_cleanup_obsolete_reviewer_comment_lease`, whose assessment paths are
+  likewise read-gated, so an operator can diagnose a stuck lease from whichever
+  namespace happens to be attached without switching roles. The dry run
+  performs no mutation and records append-only evidence in-session.
+- **`apply=true` (mutation) requires all of the following**, in order: the
+  session must have resolved exactly `cleanup_post_merge_moot_lease` (resolving
+  any other task — including a sibling reconciler task — does not authorize
+  it); the active role must be `reconciler`; the profile must hold
+  `gitea.pr.comment`; the explicit `org`/`repo` must agree with the canonical
+  repository identity, which is derived from the session binding and can never
+  be overridden by request parameters; and matching dry-run evidence must show
+  `lease_moot`, `cleanup_allowed`, and the same PR, lease session, candidate
+  head and lease marker id that are live at apply time.
+
+Everything else fails closed: a live lease on an open PR, an already-terminal
+(idempotent) lease, a lease superseded between the dry run and the apply, a
+malformed lease missing session/head/marker, and any foreign-repository target.
+The cleanup only ever appends a terminal `phase: released` marker
+(`blocker: post-merge-moot`) — it never edits or deletes another session's
+comment, and it never merges or adopts a lease.
+
 Launch a static `gitea-reconciler` MCP namespace with
 `GITEA_MCP_PROFILE=prgs-reconciler`. Profile shape is validated by
 `reconciler_profile.assess_reconciler_profile` (#304). Use the

docs/llm-workflow-runbooks.md

@@ -48,6 +48,16 @@ It extracts the issue-first, isolated-worktree, no-self-review, profile-safety,
 merge-cleanup, fail-closed, and recovery rules into a reusable package that can
 be adapted to other repositories.
 
+### Sanctioned first mutation: `create_issue` from clean control (#749)
+
+Creating a tracking issue has no issue number yet, so no `branches/issue-<N>-*`
+worktree can exist. The sanctioned path is: clean canonical control checkout
+(accepted base branch, base-equivalent to live master, no tracked dirt) →
+resolve exact `create_issue` → `gitea_create_issue`. After the issue exists,
+all further author mutations require a registered issue-backed worktree and
+lock. Do not improvise with dummy directories, borrowed worktrees, or pre-issue
+worktrees. See `skills/llm-project-workflow/workflows/create-issue.md` §18a.
+
 ## Principle: the profile is the role, not the LLM
 
 ```text

gitea_mcp_server.py

@@ -817,7 +817,11 @@ def _enforce_canonical_repository_root(
         )
 
 
-def _enforce_branches_only_author_mutation(worktree_path: str | None = None) -> None:
+def _enforce_branches_only_author_mutation(
+    worktree_path: str | None = None,
+    *,
+    task: str | None = None,
+) -> None:
     """#274: author file/branch mutations must run from a branches/ worktree.
 
     Reviewer, merger, and reconciler roles are exempt: reconciler ``close_pr``
@@ -831,6 +835,12 @@ def _enforce_branches_only_author_mutation(worktree_path: str | None = None) ->
     genuine reconciler as an author and defeat this exemption. Keying off the
     real profile role as well preserves the exemption without weakening author
     blocking — an actual author profile classifies as ``author`` in both.
+
+    #749: ``create_issue`` alone may proceed from a **clean** canonical control
+    checkout (pure remote mutation; no issue number exists yet for an
+    issue-backed worktree). Dirty roots, non-base branches, base races, and
+    every other author mutation keep the ordinary branches-only fail-closed
+    behaviour.
     """
     if (
         _effective_workspace_role() in nwb.NON_AUTHOR_ROLES
@@ -845,10 +855,35 @@ def _enforce_branches_only_author_mutation(worktree_path: str | None = None) ->
         project_root=ctx["canonical_repo_root"],
         current_branch=git_state.get("current_branch"),
     )
-    if assessment["block"]:
+    if not assessment["block"]:
-        raise RuntimeError(
+        return
-            author_mutation_worktree.format_author_mutation_worktree_error(assessment)
+
+    # #749 create_issue bootstrap: narrow phase exemption only.
+    import create_issue_bootstrap as _cib
+
+    remote_master_sha = None
+    try:
+        remote_master_sha = root_checkout_guard.resolve_remote_master_sha(
+            ctx["canonical_repo_root"]
         )
+    except Exception:
+        remote_master_sha = None
+    bootstrap = _cib.assess_create_issue_bootstrap(
+        workspace_path=workspace,
+        canonical_repo_root=ctx["canonical_repo_root"],
+        current_branch=git_state.get("current_branch"),
+        head_sha=git_state.get("head_sha"),
+        porcelain_status=git_state.get("porcelain_status") or "",
+        remote_master_sha=remote_master_sha,
+        task=task,
+    )
+    if bootstrap.get("allowed"):
+        return
+    if bootstrap.get("block") and not bootstrap.get("not_applicable"):
+        raise RuntimeError(_cib.format_create_issue_bootstrap_error(bootstrap))
+    raise RuntimeError(
+        author_mutation_worktree.format_author_mutation_worktree_error(assessment)
+    )
 
 
 def _anti_stomp_in_test_mode() -> bool:
@@ -1213,7 +1248,7 @@ def verify_preflight_purity(
         # Historical path: root + branches after purity-order when dirty paths live.
         _enforce_canonical_repository_root(worktree_path, remote=remote)
         _enforce_root_checkout_guard(worktree_path)
-        _enforce_branches_only_author_mutation(worktree_path)
+        _enforce_branches_only_author_mutation(worktree_path, task=task)
         _enforce_issue_scope_guard(
             worktree_path,
             task=task,
@@ -1247,7 +1282,7 @@ def verify_preflight_purity(
     if production_active:
         _enforce_canonical_repository_root(worktree_path, remote=remote)
         _enforce_root_checkout_guard(worktree_path)
-        _enforce_branches_only_author_mutation(worktree_path)
+        _enforce_branches_only_author_mutation(worktree_path, task=task)
         _enforce_issue_scope_guard(
             worktree_path,
             task=task,
@@ -1366,10 +1401,15 @@ def _enforce_issue_scope_guard(
             "mark_issue",
         }
     )
+    # #749: create_issue is pre-ownership (no issue number / lock can exist yet).
+    import create_issue_bootstrap as _cib
+
+    is_create_issue = _cib.is_create_issue_task(task)
     require_lock = bool(require_author_lock) or (
         authorish
         and workflow_scope_guard.production_guards_forced()
         and role == "author"
+        and not is_create_issue
     )
     assessment = workflow_scope_guard.assess_production_mutation_guards(
         workspace_path=workspace,
@@ -1381,6 +1421,7 @@ def _enforce_issue_scope_guard(
         role_kind=role,
         require_author_lock=require_lock,
         in_test_mode=_preflight_in_test_mode(),
+        mutation_task=task,
     )
     workflow_scope_guard.raise_if_blocked(assessment)
 
@@ -1394,7 +1435,11 @@ def _production_guard_block_from_exc(exc: BaseException, **extra) -> dict | None
         kind = workflow_scope_guard.BLOCKER_PRODUCTION_GUARD
         if "root_diagnostic_edit" in text or "tracked source or test edits" in text:
             kind = workflow_scope_guard.BLOCKER_ROOT_DIAGNOSTIC_EDIT
-        elif "Branches-only mutation guard" in text or "stable control checkout" in text:
+        elif (
+            "Branches-only mutation guard" in text
+            or "stable control checkout" in text
+            or "Create-issue bootstrap guard (#749)" in text
+        ):
             kind = workflow_scope_guard.BLOCKER_MISSING_WORKTREE
         elif "out-of-scope" in text or "locked to issue" in text:
             kind = workflow_scope_guard.BLOCKER_OUT_OF_SCOPE_ISSUE
@@ -1405,10 +1450,15 @@ def _production_guard_block_from_exc(exc: BaseException, **extra) -> dict | None
             reasons=[text],
             **extra,
         )
-    if "Branches-only mutation guard" in text:
+    if "Branches-only mutation guard" in text or "Create-issue bootstrap guard (#749)" in text:
+        # Prefer bootstrap next-action text when present (#749 AC4).
+        next_action = None
+        if "exact_next_action:" in text:
+            next_action = text.split("exact_next_action:", 1)[1].strip()
         return workflow_scope_guard.block_response(
             blocker_kind=workflow_scope_guard.BLOCKER_MISSING_WORKTREE,
             reasons=[text],
+            exact_next_action=next_action,
             **extra,
         )
     if "Root checkout guard" in text:
@@ -1650,6 +1700,7 @@ import issue_lock_worktree  # noqa: E402
 import issue_lock_provenance  # noqa: E402
 import issue_lock_store  # noqa: E402
 import issue_lock_adoption  # noqa: E402
+import issue_lock_recovery  # noqa: E402
 import stacked_pr_support  # noqa: E402
 import merge_approval_gate  # noqa: E402
 import review_quarantine  # noqa: E402  # #695 contaminated formal-review quarantine
@@ -1829,6 +1880,7 @@ def _seed_session_context(
 import issue_work_duplicate_gate  # noqa: E402
 import issue_workflow_labels  # noqa: E402
 import reviewer_pr_lease  # noqa: E402
+import post_merge_moot_lease_gate  # noqa: E402  # #745 reconciler cleanup gate
 import merger_lease_adoption  # noqa: E402
 import merged_cleanup_reconcile  # noqa: E402
 import worktree_cleanup_audit  # noqa: E402
@@ -2080,6 +2132,7 @@ def _assess_issue_duplicate_gate(
     auth: str,
     locked_branch: str | None = None,
     phase: str,
+    recovered_owning_pr: dict | None = None,
 ) -> dict:
     open_prs, branch_names, claim_entry = _collect_issue_duplicate_context(
         h, o, r, auth, issue_number
@@ -2091,6 +2144,7 @@ def _assess_issue_duplicate_gate(
         claim_entry=claim_entry,
         locked_branch=locked_branch,
         phase=phase,
+        recovered_owning_pr=recovered_owning_pr,
     )
 
 
@@ -3155,8 +3209,11 @@ def gitea_lock_issue(
         worktree_path, _canonical_local_git_root()
     )
     h, o, r = _resolve(remote, host, org, repo)
+    existing_issue_lock = _load_existing_issue_lock(
+        remote=remote, org=o, repo=r, issue_number=issue_number
+    )
     active_lease_block = issue_lock_store.assess_same_issue_lease_conflict(
-        _load_existing_issue_lock(remote=remote, org=o, repo=r, issue_number=issue_number),
+        existing_issue_lock,
         issue_number=issue_number,
         branch_name=branch_name,
         worktree_path=resolved_worktree,
@@ -3198,6 +3255,83 @@ def gitea_lock_issue(
         org=org,
         repo=repo,
     )
+    # ── Dead-session lock recovery assessment (#753) ──
+    # When the MCP session that took a lock exits, the lock goes non-live
+    # (stale by dead PID) even inside its lease TTL, and the branch it owns is
+    # ahead of its base by construction — so the base-equivalence gate below
+    # makes normal re-lock unreachable for every PR that already exists.
+    #
+    # This grants a waiver ONLY for that case, proven against the durable lock
+    # record plus live git/Gitea observation. A refused assessment never raises:
+    # it simply withholds the waiver, leaving the pre-existing guard to fail
+    # closed exactly as before. Recovery can only ever add permission.
+    recovery_assessment: dict | None = None
+    if (
+        existing_issue_lock
+        and existing_issue_lock.get("issue_number") == issue_number
+        and not issue_lock_store.is_lease_live(existing_issue_lock)
+    ):
+        recovery_auth = _auth(h)
+        try:
+            recovery_branches = api_get_all(
+                f"{repo_api_url(h, o, r)}/branches", recovery_auth
+            )
+        except Exception as exc:
+            raise RuntimeError(
+                f"Could not list branches to verify issue-lock recovery: {exc}"
+            )
+        recovery_remote_head: str | None = None
+        recovery_candidates: list[str] = []
+        for entry in recovery_branches:
+            entry_name = _branch_entry_name(entry)
+            if entry_name == branch_name:
+                recovery_remote_head = _branch_entry_commit_sha(entry)
+            if issue_lock_adoption.branch_carries_issue_marker(entry_name, issue_number):
+                recovery_candidates.append(entry_name)
+        recovery_pr_head: str | None = None
+        recovery_pr_number: int | None = None
+        for pull in _list_open_pulls(h, o, r, recovery_auth):
+            pull_head = pull.get("head") or {}
+            if str(pull_head.get("ref") or "") == branch_name:
+                recovery_pr_head = pull_head.get("sha")
+                recovery_pr_number = pull.get("number")
+                break
+        recovery_claimant = _work_lease_claimant(h)
+        recovery_assessment = issue_lock_recovery.assess_dead_session_lock_recovery(
+            existing_issue_lock,
+            issue_number=issue_number,
+            branch_name=branch_name,
+            worktree_path=resolved_worktree,
+            remote=remote,
+            org=o,
+            repo=r,
+            identity=recovery_claimant.get("username"),
+            profile=recovery_claimant.get("profile"),
+            current_branch=git_state.get("current_branch"),
+            porcelain_status=git_state.get("porcelain_status") or "",
+            head_sha=git_state.get("head_sha"),
+            remote_head_sha=recovery_remote_head,
+            pr_head_sha=recovery_pr_head,
+            pr_number=recovery_pr_number,
+            competing_live_locks=issue_lock_store.list_live_locks(),
+            candidate_branches=recovery_candidates,
+            current_pid=os.getpid(),
+        )
+
+    recovery_sanctioned = bool(
+        recovery_assessment and recovery_assessment.get("recovery_sanctioned")
+    )
+    # #755: a sanctioned dead-session recovery always has an owning open PR —
+    # that is what makes it a recovery rather than a fresh claim. Carry the
+    # server-derived owning-PR evidence into the duplicate-work gate below so
+    # the PR this lock already owns is not mistaken for competing duplicate
+    # work. Withheld (None) unless recovery was granted, so the ordinary
+    # duplicate blocker is untouched on every other path.
+    recovered_owning_pr = (
+        issue_lock_recovery.owning_pr_recovery_evidence(recovery_assessment)
+        if recovery_sanctioned
+        else None
+    )
     lock_assessment = issue_lock_worktree.assess_issue_lock_worktree(
         worktree_path=resolved_worktree,
         current_branch=git_state.get("current_branch"),
@@ -3205,10 +3339,20 @@ def gitea_lock_issue(
         base_equivalent=git_state.get("base_equivalent"),
         inspected_git_root=git_state.get("inspected_git_root"),
         base_branch=git_state.get("base_branch"),
+        recovery_sanctioned=recovery_sanctioned,
     )
     if lock_assessment["block"]:
+        reasons = list(lock_assessment.get("reasons") or [])
+        # Surface why recovery was unavailable, so a blocked caller sees the
+        # exact missing evidence instead of only the base-equivalence text.
+        if recovery_assessment and recovery_assessment.get("is_candidate"):
+            reasons.append(
+                issue_lock_recovery.format_recovery_refusal(recovery_assessment)
+            )
         raise RuntimeError(
-            issue_lock_worktree.format_issue_lock_worktree_error(lock_assessment)
+            issue_lock_worktree.format_issue_lock_worktree_error(
+                {**lock_assessment, "reasons": reasons}
+            )
         )
 
     auth = _auth(h)
@@ -3220,6 +3364,7 @@ def gitea_lock_issue(
         auth=auth,
         locked_branch=branch_name,
         phase=issue_work_duplicate_gate.PHASE_LOCK,
+        recovered_owning_pr=recovered_owning_pr,
     )
     if duplicate_gate.get("block"):
         raise ValueError("; ".join(duplicate_gate.get("reasons") or [
@@ -3271,6 +3416,14 @@ def gitea_lock_issue(
     }
     if stacked_approved:
         data["approved_stacked_base"] = stacked_approved
+    if recovery_sanctioned and recovery_assessment:
+        # #753 AC2/AC6: record that this claim was recovered after session
+        # death, with the prior and replacement session identity, so the
+        # takeover is auditable and never looks like an original claim.
+        data["dead_session_recovery"] = issue_lock_recovery.build_recovery_record(
+            recovery_assessment,
+            recovered_at=_work_lease_timestamp(_work_lease_now()),
+        )
 
     lock_file_path = _save_issue_lock(data)
     lock_record = issue_lock_store.read_lock_file(lock_file_path) or data
@@ -3304,6 +3457,14 @@ def gitea_lock_issue(
         "lock_freshness": freshness,
         "lock_proof": lock_proof,
     }
+    if recovery_sanctioned and recovery_assessment:
+        result["dead_session_recovery"] = data["dead_session_recovery"]
+        result["message"] = (
+            f"Recovered the durable lock for issue #{issue_number} on branch "
+            f"'{branch_name}' after the owning MCP session (pid "
+            f"{recovery_assessment['evidence'].get('prior_session_pid')}) exited; "
+            "ownership evidence matched exactly (fail-closed check complete)."
+        )
     if stacked_approved:
         result["approved_stacked_base"] = stacked_approved
         result["message"] = (
@@ -8859,13 +9020,14 @@ def gitea_review_pr(
         return out
 
 
-def _delete_branch_repository_binding_block(
+def _repository_binding_block(
     remote: str | None,
     *,
     org: str | None,
     repo: str | None,
+    required_permission: str = "gitea.branch.delete",
 ) -> dict | None:
-    """#733: validate an explicit delete target against the workspace binding.
+    """#733: validate an explicit mutation target against the workspace binding.
 
     The trusted repository identity comes only from the verified,
     workspace-aligned git remote (never ``REMOTES`` defaults, never
@@ -8898,7 +9060,7 @@ def _delete_branch_repository_binding_block(
         return {
             "success": False,
             "performed": False,
-            "required_permission": "gitea.branch.delete",
+            "required_permission": required_permission,
             "reasons": canonical_reasons,
             "blocker_kind": "repository_binding",
         }
@@ -8917,9 +9079,9 @@ def _delete_branch_repository_binding_block(
             return {
                 "success": False,
                 "performed": False,
-                "required_permission": "gitea.branch.delete",
+                "required_permission": required_permission,
                 "reasons": list(override.get("reasons") or [
-                    "delete target repository does not match the workspace "
+                    "target repository does not match the workspace "
                     "binding (fail closed)"
                 ]),
                 "blocker_kind": "repository_binding",
@@ -8929,17 +9091,51 @@ def _delete_branch_repository_binding_block(
         return {
             "success": False,
             "performed": False,
-            "required_permission": "gitea.branch.delete",
+            "required_permission": required_permission,
             "reasons": [
                 "repository binding unverified: no workspace repository "
                 "identity could be established to corroborate the explicit "
-                "delete target (fail closed)"
+                "target (fail closed)"
             ],
             "blocker_kind": "repository_binding",
         }
     return None
 
 
+def _delete_branch_repository_binding_block(
+    remote: str | None,
+    *,
+    org: str | None,
+    repo: str | None,
+) -> dict | None:
+    """Delete-branch view of :func:`_repository_binding_block` (#733).
+
+    Retained as the named entry point for the delete path so #733/#739
+    regression coverage keeps exercising the exact permission label that
+    ``gitea_delete_branch`` reports.
+    """
+    return _repository_binding_block(
+        remote, org=org, repo=repo,
+        required_permission="gitea.branch.delete",
+    )
+
+
+def _bound_repository_slug(remote: str | None) -> str | None:
+    """Canonical repository slug for the active session, or None (#745).
+
+    Same trust order as ``_repository_binding_block``: the configured canonical
+    root when the namespace declares one, otherwise the workspace-derived
+    identity. Request parameters are never consulted, so they cannot redirect a
+    mutation at a foreign repository.
+    """
+    canonical_slug, canonical_reasons = _canonical_repository_slug(
+        get_profile(), remote
+    )
+    if canonical_reasons:
+        return None
+    return canonical_slug or _workspace_repository_slug(remote)
+
+
 @mcp.tool()
 def gitea_delete_branch(
     branch: str,
@@ -12368,6 +12564,10 @@ def gitea_diagnose_reviewer_pr_lease_handoff(
 def gitea_cleanup_post_merge_moot_lease(
     pr_number: int,
     apply: bool = False,
+    expected_session_id: str | None = None,
+    expected_candidate_head: str | None = None,
+    expected_lease_comment_id: int | None = None,
+    worktree_path: str | None = None,
     remote: str = "dadeschools",
     host: str | None = None,
     org: str | None = None,
@@ -12383,14 +12583,37 @@ def gitea_cleanup_post_merge_moot_lease(
     an append-only comment that neutralises the moot lease without deleting any
     other session's comment.
 
+    Role binding (#745). The two modes are gated differently, on purpose:
+
+    * ``apply=false`` (assessment) stays reachable under ``gitea.read`` for
+      **any** role, matching ``gitea_cleanup_stale_review_decision_lock`` and
+      ``gitea_cleanup_obsolete_reviewer_comment_lease``, so an operator can
+      diagnose a stuck lease from whichever namespace is attached. It performs
+      no mutation and records append-only dry-run evidence in-session.
+    * ``apply=true`` (mutation) additionally requires, in order: the session to
+      have resolved exactly ``cleanup_post_merge_moot_lease``; the
+      ``reconciler`` role; ``gitea.pr.comment``; a validated repository /
+      canonical-root / workspace binding; and matching dry-run evidence proving
+      ``lease_moot`` and ``cleanup_allowed`` for the same PR, lease session,
+      candidate head and lease marker. Author, reviewer and merger fail closed
+      here even though their profiles carry ``gitea.pr.comment``.
+
     Args:
         pr_number: The PR whose lingering lease to assess/clean.
         apply: When false (default) report only (read-only). When true, post the
-            terminal released marker if — and only if — cleanup is allowed.
+            terminal released marker if — and only if — cleanup is authorized.
+        expected_session_id: Optional lease session the caller expects; a
+            mismatch against the live lease fails closed.
+        expected_candidate_head: Optional leased head the caller expects; a
+            mismatch against the live lease fails closed.
+        expected_lease_comment_id: Optional lease marker id the caller expects;
+            a mismatch against the live lease fails closed.
+        worktree_path: Reconciler worktree to bind the apply-path preflight to.
         remote: Known instance — 'dadeschools' or 'prgs'.
         host: Override the Gitea host.
-        org: Override the owner/organization.
+        org: Override the owner/organization. Validated against the canonical
-        repo: Override the repository name.
+            repository identity; it can never redirect the mutation.
+        repo: Override the repository name. Validated as above.
 
     Returns:
         dict reporting PR merged/closed state, merge_commit_sha, linked-issue
@@ -12450,14 +12673,62 @@ def gitea_cleanup_post_merge_moot_lease(
         "reasons": assessment.get("reasons") or [],
     }
 
+    # The canonical repository identity comes from the session binding only —
+    # never from the org/repo request parameters (#745 requirement 10).
+    repository_slug = _bound_repository_slug(remote)
+    report["repository_slug"] = repository_slug
+    report["required_task"] = post_merge_moot_lease_gate.CLEANUP_TASK
+    report["required_role_kind"] = post_merge_moot_lease_gate.REQUIRED_ROLE
+
     if not apply:
+        # Append-only dry-run evidence. Recorded for every assessment, allowed
+        # or not, so a later apply can prove the lease it saw is the lease that
+        # is still there. Prior entries are never rewritten.
+        evidence = post_merge_moot_lease_gate.record_dry_run(
+            pr_number=pr_number,
+            repository_slug=repository_slug,
+            lease_moot=bool(assessment.get("is_moot")),
+            cleanup_allowed=bool(assessment.get("cleanup_allowed")),
+            session_id=active.get("session_id"),
+            candidate_head=active.get("candidate_head"),
+            lease_comment_id=active.get("comment_id"),
+        )
+        report["dry_run_evidence"] = evidence
         return report
+
+    # ---------------------------------------------------------------- apply --
+    # Preserve the existing dry-run safety assessment: a lease that is not moot
+    # (open PR, already-terminal, absent) is refused here exactly as before, and
+    # no mutation is attempted.
     if not assessment.get("cleanup_allowed"):
         report["cleanup_skipped_reason"] = (
             assessment.get("reasons") or ["cleanup not allowed"]
         )
         return report
 
+    # 1 + 2 + 6: exact resolved cleanup task, reconciler role, and matching
+    # append-only dry-run evidence. Permission alone is deliberately not enough.
+    authorization = post_merge_moot_lease_gate.assess_apply_authorization(
+        pr_number=pr_number,
+        repository_slug=repository_slug,
+        resolved_task=_preflight_resolved_task,
+        active_role_kind=_profile_role_kind(get_profile()),
+        assessment=assessment,
+        evidence=post_merge_moot_lease_gate.latest_dry_run(
+            pr_number=pr_number, repository_slug=repository_slug
+        ),
+        expected_session_id=expected_session_id,
+        expected_candidate_head=expected_candidate_head,
+        expected_lease_comment_id=expected_lease_comment_id,
+    )
+    report["authorization"] = authorization
+    if not authorization["allowed"]:
+        report["success"] = False
+        report["reasons"] = authorization["reasons"]
+        report["blocker_kind"] = authorization["blocker_kind"]
+        return report
+
+    # 3: the dedicated mutation permission.
     comment_block = _profile_operation_gate("gitea.pr.comment")
     if comment_block:
         report["success"] = False
@@ -12465,7 +12736,28 @@ def gitea_cleanup_post_merge_moot_lease(
         report["permission_report"] = _permission_block_report("gitea.pr.comment")
         return report
 
-    verify_preflight_purity(remote)
+    # 4: explicit target must agree with the canonical repository identity
+    # before any preflight or mutation (#733/#739).
+    repo_binding_block = _repository_binding_block(
+        remote, org=org, repo=repo,
+        required_permission="gitea.pr.comment",
+    )
+    if repo_binding_block:
+        report["success"] = False
+        report["reasons"] = repo_binding_block["reasons"]
+        report["blocker_kind"] = repo_binding_block["blocker_kind"]
+        return report
+
+    # 4 (cont.): canonical root, workspace binding and the resolved-task match
+    # are enforced by the shared preflight, with the explicit target forwarded
+    # so the #604 anti-stomp resolution validates the targeted repository.
+    verify_preflight_purity(
+        remote,
+        worktree_path=worktree_path,
+        task=post_merge_moot_lease_gate.CLEANUP_TASK,
+        org=org,
+        repo=repo,
+    )
     body = assessment["release_body"]
     comment_url = f"{repo_api_url(h, o, r)}/issues/{pr_number}/comments"
     with _audited(
@@ -15367,46 +15659,151 @@ def _count_commits_behind(
     return None
 
 
+def _matching_protection_rule(protections: Any, branch: str) -> dict | None:
+    """Select the branch-protection rule governing ``branch`` (#751)."""
+    if not isinstance(protections, list):
+        return None
+    for rule in protections:
+        if not isinstance(rule, dict):
+            continue
+        name = (rule.get("branch_name") or rule.get("rule_name") or "").strip()
+        # Exact match or glob-ish contains for common patterns.
+        if name == branch or name in ("*", f"{branch}"):
+            return rule
+    # Fallback: any protection that mentions the base branch.
+    for rule in protections:
+        if not isinstance(rule, dict):
+            continue
+        name = (rule.get("branch_name") or rule.get("rule_name") or "").strip()
+        if name and (branch in name or name.endswith(branch)):
+            return rule
+    return None
+
+
+def _branch_protection_policy(
+    base_url: str,
+    auth: dict,
+    *,
+    base_branch: str,
+) -> dict:
+    """Read the live branch-protection policy for ``base_branch`` (#751).
+
+    Exposes both the current-base rule (``block_on_outdated_branch``) and the
+    status-check requirement (``enable_status_check`` / ``status_check_contexts``)
+    from the same payload, so the checks assessor can tell "no checks required"
+    apart from "required checks pending".
+
+    ``determinable`` is False only when the policy genuinely could not be read
+    (missing branch or API failure) — never merely because no rule exists. A
+    successful read that finds no protection for the branch is authoritative
+    evidence that status checks are not required.
+    """
+    policy: dict[str, Any] = {
+        "determinable": False,
+        "protection_found": False,
+        "requires_current_base": None,
+        "checks_enabled": None,
+        "required_contexts": [],
+        "base_branch": (base_branch or "").strip() or None,
+    }
+    branch = (base_branch or "").strip()
+    if not branch:
+        return policy
+
+    def _apply(rule: dict) -> None:
+        policy["protection_found"] = True
+        if "block_on_outdated_branch" in rule:
+            policy["requires_current_base"] = bool(
+                rule.get("block_on_outdated_branch")
+            )
+        if "enable_status_check" in rule:
+            policy["checks_enabled"] = bool(rule.get("enable_status_check"))
+        contexts = rule.get("status_check_contexts")
+        if isinstance(contexts, list):
+            policy["required_contexts"] = [
+                str(ctx).strip() for ctx in contexts if str(ctx or "").strip()
+            ]
+
+    try:
+        protections = api_request(
+            "GET", f"{base_url}/branch_protections", auth
+        )
+        rule = _matching_protection_rule(protections, branch)
+        if rule is not None:
+            _apply(rule)
+        if not policy["protection_found"]:
+            # Branch payload may embed effective protection.
+            br = api_request("GET", f"{base_url}/branches/{branch}", auth) or {}
+            prot = br.get("protection") or br.get("effective_branch_protection") or {}
+            if isinstance(prot, dict) and prot:
+                _apply(prot)
+        policy["determinable"] = True
+    except Exception:
+        # Genuine read failure — leave determinable False so callers fail closed.
+        return policy
+
+    if not policy["protection_found"]:
+        # Authoritative absence: no protection governs the branch, so no status
+        # check is required by policy.
+        policy["checks_enabled"] = False
+    elif policy["checks_enabled"] is None:
+        # Protection exists but omits the status-check field entirely: Gitea
+        # only enforces contexts when the flag is set, so absence means off.
+        policy["checks_enabled"] = False
+
+    return policy
+
+
 def _branch_protection_requires_current_base(
     base_url: str,
     auth: dict,
     *,
     base_branch: str,
 ) -> bool | None:
-    """Read Gitea branch protection ``block_on_outdated_branch`` when present."""
+    """Read Gitea branch protection ``block_on_outdated_branch`` when present.
-    branch = (base_branch or "").strip()
+
-    if not branch:
+    Thin accessor over :func:`_branch_protection_policy`; return semantics are
-        return None
+    unchanged (``None`` when the rule is absent or unreadable).
+    """
+    policy = _branch_protection_policy(base_url, auth, base_branch=base_branch)
+    return policy.get("requires_current_base")
+
+
+def _commit_checks_snapshot(
+    base_url: str,
+    auth: dict,
+    *,
+    sha: str,
+) -> dict:
+    """Read the combined commit status *and its context collection* (#751).
+
+    The combined ``state`` alone is not decisive: Gitea reports ``pending`` for
+    a commit with an empty status-context collection, which is indistinguishable
+    from executing CI unless the collection itself is inspected.
+    """
+    snapshot: dict[str, Any] = {
+        "determinable": False,
+        "combined_state": None,
+        "statuses": [],
+    }
+    head = (sha or "").strip()
+    if not head:
+        return snapshot
     try:
-        # Prefer the named protection rule matching the base branch.
+        payload = api_request(
-        protections = api_request(
+            "GET", f"{base_url}/commits/{head}/status", auth
-            "GET", f"{base_url}/branch_protections", auth
+        )
-        ) or []
-        if isinstance(protections, list):
-            for rule in protections:
-                if not isinstance(rule, dict):
-                    continue
-                name = (rule.get("branch_name") or rule.get("rule_name") or "").strip()
-                # Exact match or glob-ish contains for common patterns.
-                if name == branch or name in ("*", f"{branch}"):
-                    if "block_on_outdated_branch" in rule:
-                        return bool(rule.get("block_on_outdated_branch"))
-            # Fallback: any protection that mentions the base branch.
-            for rule in protections:
-                if not isinstance(rule, dict):
-                    continue
-                name = (rule.get("branch_name") or rule.get("rule_name") or "").strip()
-                if branch in name or name.endswith(branch):
-                    if "block_on_outdated_branch" in rule:
-                        return bool(rule.get("block_on_outdated_branch"))
-        # Branch payload may embed effective protection.
-        br = api_request("GET", f"{base_url}/branches/{branch}", auth) or {}
-        prot = br.get("protection") or br.get("effective_branch_protection") or {}
-        if isinstance(prot, dict) and "block_on_outdated_branch" in prot:
-            return bool(prot.get("block_on_outdated_branch"))
     except Exception:
-        return None
+        return snapshot
-    return None
+    if payload is None:
+        return snapshot
+    snapshot["determinable"] = True
+    if not isinstance(payload, dict):
+        return snapshot
+    snapshot["combined_state"] = (payload.get("state") or "").strip().lower() or None
+    statuses = payload.get("statuses")
+    snapshot["statuses"] = statuses if isinstance(statuses, list) else []
+    return snapshot
 
 
 def _prove_author_ownership_for_pr(
@@ -15651,11 +16048,14 @@ def gitea_assess_pr_sync_status(
             # Fail closed on unknown behind-count only when SHAs differ.
             commits_behind = 0 if pr_head_sha == base_head_sha else None
 
+    # Single live read of the base-branch protection policy; it carries both
+    # the current-base rule and the status-check requirement (#751).
+    protection_policy = _branch_protection_policy(
+        base, auth, base_branch=base_branch
+    )
     if branch_protection_requires_current_base is None:
-        branch_protection_requires_current_base = (
+        branch_protection_requires_current_base = protection_policy.get(
-            _branch_protection_requires_current_base(
+            "requires_current_base"
-                base, auth, base_branch=base_branch
-            )
         )
     # When protection cannot be read, fail closed by requiring current base
     # whenever the PR is behind (safer for protected repos). Callers may pass
@@ -15722,23 +16122,30 @@ def gitea_assess_pr_sync_status(
     except Exception:
         pass
 
+    # ── Live checks derivation (#751) ────────────────────────────────────
+    # Classify from the actual status-context collection plus the live
+    # protection policy. The combined ``state`` is never treated as proof that
+    # CI is executing, because Gitea reports ``pending`` for an empty
+    # collection. ``checks_required`` is always derived from live evidence and
+    # is deliberately not a caller-supplied input, so no session can declare
+    # checks optional without proof.
+    checks_snapshot = _commit_checks_snapshot(base, auth, sha=pr_head_sha or "")
+    checks_classification = pr_sync_status.classify_commit_checks(
+        combined_state=checks_snapshot.get("combined_state"),
+        statuses=checks_snapshot.get("statuses"),
+        checks_enabled=protection_policy.get("checks_enabled"),
+        required_contexts=protection_policy.get("required_contexts"),
+        policy_determinable=bool(protection_policy.get("determinable")),
+        status_determinable=bool(checks_snapshot.get("determinable")),
+    )
+    checks_required = bool(checks_classification.get("checks_required", True))
+    caller_supplied_checks_status = checks_status
     if checks_status is None:
-        checks_status = "unknown"
+        checks_status = checks_classification.get("checks_status")
-        # Combined status on PR head when Actions/status API is available.
+    elif checks_classification.get("checks_status") != pr_sync_status.CHECKS_UNKNOWN:
-        if pr_head_sha:
+        # Live evidence outranks a caller-supplied value; the override only
-            try:
+        # applies when live status could not be classified at all.
-                st = api_request(
+        checks_status = checks_classification.get("checks_status")
-                    "GET",
-                    f"{base}/commits/{pr_head_sha}/status",
-                    auth,
-                ) or {}
-                state = (st.get("state") or "").strip().lower()
-                if state:
-                    checks_status = state
-                elif not st:
-                    checks_status = "none"
-            except Exception:
-                checks_status = "unknown"
 
     assessment = pr_sync_status.assess_pr_sync_status(
         host=h,
@@ -15759,9 +16166,26 @@ def gitea_assess_pr_sync_status(
         active_reviewer_lease=active_reviewer_lease,
         active_merger_lease=active_merger_lease,
         prepared_verdict_head_sha=prepared_verdict_head_sha,
+        checks_required=checks_required,
     )
     assessment["remote"] = remote if remote in REMOTES else None
     assessment["base_branch"] = base_branch
+    # Evidence for the checks decision (#751) — no secrets, read-only.
+    assessment["checks_evidence"] = {
+        "combined_state": checks_classification.get("combined_state"),
+        "context_count": checks_classification.get("context_count"),
+        "observed_contexts": checks_classification.get("observed_contexts"),
+        "required_contexts": checks_classification.get("required_contexts"),
+        "missing_required_contexts": checks_classification.get(
+            "missing_required_contexts"
+        ),
+        "policy_determinable": checks_classification.get("policy_determinable"),
+        "status_determinable": checks_classification.get("status_determinable"),
+        "protection_found": protection_policy.get("protection_found"),
+        "checks_enabled": protection_policy.get("checks_enabled"),
+        "caller_supplied_checks_status": caller_supplied_checks_status,
+        "reasons": checks_classification.get("reasons"),
+    }
     assessment["success"] = True
     assessment["performed"] = False
     return assessment

issue_lock_adoption.py

@@ -85,6 +85,15 @@ def _branch_carries_issue_marker(branch_name: str, issue_number: int) -> bool:
     return re.search(pattern, name) is not None
 
 
+def branch_carries_issue_marker(branch_name: str, issue_number: int) -> bool:
+    """Public accessor for the exact issue-marker match (#753).
+
+    Dead-session lock recovery needs the same word-boundary matcher to detect
+    ambiguous branch claims, so it is exposed rather than reached into.
+    """
+    return _branch_carries_issue_marker(branch_name, issue_number)
+
+
 def assess_own_branch_adoption(
     *,
     issue_number: int,

issue_lock_recovery.py

@@ -0,0 +1,451 @@
+"""Dead-session author issue-lock recovery (#753).
+
+A durable author issue lock records the PID of the MCP session that took it.
+When that process exits, ``issue_lock_store.assess_lock_freshness`` classifies
+the lock as ``stale`` (``live=False``) even while its lease is still within TTL,
+so every ownership check that requires a *live* lock fails closed.
+
+Re-taking the lock through ``gitea_lock_issue`` is unreachable for real work:
+``issue_lock_worktree.assess_issue_lock_worktree`` demands the worktree be
+base-equivalent to ``master``/``main``/``dev``, and a branch that already
+carries commits is ahead of its base by construction. The existing
+``assess_expired_lock_reclaim`` affordance does not apply either, because
+``assess_same_issue_lease_conflict`` only consults it once the lease has
+*expired* — a dead PID under an unexpired lease never reaches it.
+
+This module is the pure evidence assessor for that one narrow case. It grants
+recovery only when every element of durable ownership still matches exactly and
+the recorded process is demonstrably dead. It never trusts caller assertions:
+every field is compared against durable lock state or live observation supplied
+by the caller. It performs no mutation and no network I/O.
+
+Recovery deliberately does **not** relax base-equivalence for brand-new issue
+claims — only for a lock whose own prior record already proves the branch,
+worktree, head, and author.
+"""
+
+from __future__ import annotations
+
+import os
+from typing import Any, Iterable, Mapping, Sequence
+
+from issue_lock_store import is_process_alive
+from reviewer_worktree import parse_dirty_tracked_files
+
+# Outcome values
+RECOVERY_SANCTIONED = "RECOVERY_SANCTIONED"
+NO_CANDIDATE = "NO_CANDIDATE"
+REFUSED = "REFUSED"
+
+# Durable fields a lock must carry before it can be considered at all.
+REQUIRED_LOCK_FIELDS = ("issue_number", "branch_name", "worktree_path")
+
+
+def _same_realpath(left: str | None, right: str | None) -> bool:
+    if not left or not right:
+        return False
+    try:
+        return os.path.realpath(left) == os.path.realpath(right)
+    except OSError:
+        return left == right
+
+
+def _text(value: Any) -> str:
+    return str(value or "").strip()
+
+
+def _lock_claimant(lock: Mapping[str, Any]) -> dict[str, Any]:
+    claimant = lock.get("claimant")
+    if not isinstance(claimant, Mapping):
+        lease = lock.get("work_lease")
+        claimant = lease.get("claimant") if isinstance(lease, Mapping) else None
+    return dict(claimant) if isinstance(claimant, Mapping) else {}
+
+
+def _recorded_pid(lock: Mapping[str, Any]) -> Any:
+    pid = lock.get("session_pid")
+    if pid is None:
+        pid = lock.get("pid")
+    return pid
+
+
+def _malformed_reasons(lock: Mapping[str, Any]) -> list[str]:
+    """Names of durable fields that are missing or unusable."""
+    missing: list[str] = []
+    for field in REQUIRED_LOCK_FIELDS:
+        if not _text(lock.get(field)):
+            missing.append(field)
+    pid = _recorded_pid(lock)
+    if pid is None or _text(pid) == "":
+        missing.append("session_pid/pid")
+    else:
+        try:
+            if int(pid) <= 0:
+                missing.append("session_pid/pid")
+        except (TypeError, ValueError):
+            missing.append("session_pid/pid")
+    return missing
+
+
+def assess_dead_session_lock_recovery(
+    existing_lock: Mapping[str, Any] | None,
+    *,
+    issue_number: int,
+    branch_name: str,
+    worktree_path: str,
+    remote: str,
+    org: str,
+    repo: str,
+    identity: str | None,
+    profile: str | None,
+    current_branch: str | None,
+    porcelain_status: str,
+    head_sha: str | None,
+    remote_head_sha: str | None,
+    pr_head_sha: str | None = None,
+    pr_number: int | None = None,
+    competing_live_locks: Sequence[Mapping[str, Any]] | None = None,
+    candidate_branches: Iterable[str] | None = None,
+    current_pid: int | None = None,
+) -> dict[str, Any]:
+    """Decide whether a dead-session author lock may be natively recovered.
+
+    Returns a dict with ``recovery_sanctioned`` (bool), ``outcome``, ``reasons``
+    (why it was refused, or the positive proof when sanctioned), and
+    ``evidence`` (a redaction-safe record for auditing).
+
+    ``NO_CANDIDATE`` means no recovery was attempted at all — there is no
+    existing lock, or the lock does not describe this issue. The caller must
+    treat that exactly as it treated the pre-#753 world. ``REFUSED`` means a
+    candidate existed but the evidence did not agree; the caller fails closed.
+    """
+    reasons: list[str] = []
+    evidence: dict[str, Any] = {
+        "issue_number": issue_number,
+        "branch_name": branch_name,
+        "worktree_path": worktree_path,
+        "remote": remote,
+        "org": org,
+        "repo": repo,
+    }
+
+    if not existing_lock:
+        return _result(
+            NO_CANDIDATE, False, ["no existing durable lock for this issue"], evidence
+        )
+
+    lock = dict(existing_lock)
+
+    # ── Candidate identification ────────────────────────────────────────────
+    # Recovery only ever applies to a lock that already claims THIS issue.
+    # Anything else is not a recovery candidate and must not be reinterpreted.
+    if lock.get("issue_number") != issue_number:
+        return _result(
+            NO_CANDIDATE,
+            False,
+            [
+                f"existing lock targets issue #{lock.get('issue_number')}, "
+                f"not #{issue_number}; not a recovery candidate"
+            ],
+            evidence,
+        )
+
+    # A malformed/incomplete durable record can never prove ownership.
+    missing = _malformed_reasons(lock)
+    if missing:
+        return _result(
+            REFUSED,
+            False,
+            [
+                "durable lock record is incomplete and cannot prove ownership "
+                f"(missing/unusable: {', '.join(missing)})"
+            ],
+            evidence,
+        )
+
+    recorded_pid = _recorded_pid(lock)
+    evidence["prior_session_pid"] = recorded_pid
+    evidence["replacement_session_pid"] = (
+        current_pid if current_pid is not None else os.getpid()
+    )
+
+    # ── Repository scope ────────────────────────────────────────────────────
+    for field, expected in (("remote", remote), ("org", org), ("repo", repo)):
+        actual = _text(lock.get(field))
+        if actual != _text(expected):
+            reasons.append(
+                f"lock {field} '{actual}' does not match requested '{_text(expected)}'"
+            )
+
+    # ── Branch identity ─────────────────────────────────────────────────────
+    locked_branch = _text(lock.get("branch_name"))
+    if locked_branch != _text(branch_name):
+        reasons.append(
+            f"lock branch '{locked_branch}' does not match requested "
+            f"'{_text(branch_name)}'"
+        )
+    evidence["locked_branch"] = locked_branch
+
+    # The worktree must actually be sitting on the locked branch. Without this
+    # a clean worktree parked elsewhere could stand in for the real work.
+    checked_out = _text(current_branch)
+    if not checked_out:
+        reasons.append(
+            "worktree is not on a named branch (detached HEAD); locked-branch "
+            "occupancy could not be proven"
+        )
+    elif checked_out != locked_branch:
+        reasons.append(
+            f"worktree is on branch '{checked_out}', not the locked branch "
+            f"'{locked_branch}'"
+        )
+
+    # ── Worktree identity ───────────────────────────────────────────────────
+    locked_worktree = _text(lock.get("worktree_path"))
+    if not _same_realpath(locked_worktree, worktree_path):
+        reasons.append(
+            f"lock worktree '{locked_worktree}' does not match declared "
+            f"'{_text(worktree_path)}'"
+        )
+    evidence["locked_worktree_path"] = locked_worktree
+
+    # ── Cleanliness (never waived) ──────────────────────────────────────────
+    dirty_files = parse_dirty_tracked_files(porcelain_status)
+    if dirty_files:
+        reasons.append(
+            "worktree has tracked local edits; recovery requires a clean "
+            f"worktree (dirty files: {', '.join(dirty_files)})"
+        )
+    evidence["dirty_files"] = dirty_files
+
+    # ── Head agreement: local == remote == PR ───────────────────────────────
+    local_head = _text(head_sha)
+    remote_head = _text(remote_head_sha)
+    if not local_head:
+        reasons.append("local head SHA could not be determined")
+    if not remote_head:
+        reasons.append(
+            f"remote head for branch '{locked_branch}' could not be determined"
+        )
+    if local_head and remote_head and local_head != remote_head:
+        reasons.append(
+            f"local head {local_head} does not match remote branch head {remote_head}"
+        )
+    evidence["local_head"] = local_head or None
+    evidence["remote_head"] = remote_head or None
+
+    pr_head = _text(pr_head_sha)
+    if pr_head:
+        evidence["pr_head"] = pr_head
+        evidence["pr_number"] = pr_number
+        if local_head and pr_head != local_head:
+            reasons.append(
+                f"open PR #{pr_number} head {pr_head} does not match local head "
+                f"{local_head}"
+            )
+
+    # ── Author identity ─────────────────────────────────────────────────────
+    claimant = _lock_claimant(lock)
+    locked_identity = _text(claimant.get("username"))
+    locked_profile = _text(claimant.get("profile"))
+    evidence["locked_identity"] = locked_identity or None
+    evidence["locked_profile"] = locked_profile or None
+    if not locked_identity or not locked_profile:
+        reasons.append(
+            "durable lock does not record a claimant identity/profile; "
+            "author ownership could not be proven"
+        )
+    if not _text(identity) or not _text(profile):
+        reasons.append(
+            "active session identity/profile is unknown; author ownership "
+            "could not be proven"
+        )
+    if locked_identity and _text(identity) and locked_identity != _text(identity):
+        reasons.append(
+            f"lock claimant '{locked_identity}' does not match active identity "
+            f"'{_text(identity)}'"
+        )
+    if locked_profile and _text(profile) and locked_profile != _text(profile):
+        reasons.append(
+            f"lock profile '{locked_profile}' does not match active profile "
+            f"'{_text(profile)}'"
+        )
+
+    # ── The defining condition: the recorded owner must be dead ─────────────
+    prior_alive = is_process_alive(recorded_pid)
+    evidence["prior_pid_alive"] = prior_alive
+    if prior_alive:
+        reasons.append(
+            f"prior owner pid {recorded_pid} is still alive; this is not a "
+            "dead-session recovery"
+        )
+    if current_pid is not None and recorded_pid is not None:
+        try:
+            if int(recorded_pid) == int(current_pid):
+                reasons.append(
+                    "recorded pid is the current session; nothing to recover"
+                )
+        except (TypeError, ValueError):
+            pass
+
+    # ── Competing ownership ─────────────────────────────────────────────────
+    competing: list[dict[str, Any]] = []
+    for entry in competing_live_locks or ():
+        if not isinstance(entry, Mapping):
+            continue
+        same_issue = entry.get("issue_number") == issue_number
+        same_branch = _text(entry.get("branch_name")) == locked_branch
+        if not (same_issue or same_branch):
+            continue
+        # The lock we are recovering is not competition with itself.
+        if (
+            same_issue
+            and same_branch
+            and _same_realpath(_text(entry.get("worktree_path")), worktree_path)
+        ):
+            continue
+        competing.append(
+            {
+                "issue_number": entry.get("issue_number"),
+                "branch_name": entry.get("branch_name"),
+                "worktree_path": entry.get("worktree_path"),
+                "pid": entry.get("pid"),
+            }
+        )
+    if competing:
+        described = ", ".join(
+            f"issue #{c['issue_number']} branch '{c['branch_name']}'" for c in competing
+        )
+        reasons.append(f"competing live lock or lease exists ({described})")
+    evidence["competing_live_locks"] = competing
+
+    # ── Ambiguous branch claims ─────────────────────────────────────────────
+    others = [
+        name
+        for name in (candidate_branches or ())
+        if _text(name) and _text(name) != locked_branch
+    ]
+    if others:
+        reasons.append(
+            "multiple branches claim this issue "
+            f"({', '.join(sorted(set(others)))}); ownership is ambiguous"
+        )
+    evidence["other_candidate_branches"] = sorted(set(others))
+
+    if reasons:
+        return _result(REFUSED, False, reasons, evidence)
+
+    return _result(
+        RECOVERY_SANCTIONED,
+        True,
+        [
+            f"durable lock for issue #{issue_number} matches branch "
+            f"'{locked_branch}', worktree '{locked_worktree}', head {local_head}, "
+            f"and claimant '{locked_identity}'; recorded pid {recorded_pid} is dead"
+        ],
+        evidence,
+    )
+
+
+def _result(
+    outcome: str,
+    sanctioned: bool,
+    reasons: list[str],
+    evidence: dict[str, Any],
+) -> dict[str, Any]:
+    return {
+        "outcome": outcome,
+        "recovery_sanctioned": sanctioned,
+        "is_candidate": outcome != NO_CANDIDATE,
+        "reasons": reasons,
+        "evidence": evidence,
+    }
+
+
+def owning_pr_recovery_evidence(
+    assessment: Mapping[str, Any] | None,
+) -> dict[str, Any] | None:
+    """Server-derived proof of the open PR a sanctioned recovery already owns (#755).
+
+    A dead-session recovery is, by construction, recovery of work that already
+    has an open PR — so the duplicate-work gate's linked-open-PR blocker would
+    otherwise discard every sanctioned recovery. This distils the completed
+    assessment into the minimum evidence that gate needs to tell "the PR this
+    lock already owns" apart from "a competing duplicate PR".
+
+    Returns ``None`` unless recovery was actually granted and the assessment's
+    own evidence names exactly one owning PR whose head agrees with the local
+    and remote heads. Nothing here is caller-supplied: every field is copied
+    from evidence the assessor built out of durable lock state plus live
+    git/Gitea observation, so a caller cannot manufacture an exemption.
+    """
+    if not isinstance(assessment, Mapping):
+        return None
+    if assessment.get("outcome") != RECOVERY_SANCTIONED:
+        return None
+    if not assessment.get("recovery_sanctioned"):
+        return None
+
+    evidence = assessment.get("evidence") or {}
+    branch_name = _text(evidence.get("locked_branch"))
+    pr_head = _text(evidence.get("pr_head"))
+    local_head = _text(evidence.get("local_head"))
+    remote_head = _text(evidence.get("remote_head"))
+    raw_pr_number = evidence.get("pr_number")
+
+    if raw_pr_number is None or not branch_name or not pr_head:
+        return None
+    # The assessor already required these to agree. Re-check, so a truncated or
+    # hand-built evidence map can never authorize an exemption.
+    if pr_head != local_head or pr_head != remote_head:
+        return None
+    try:
+        pr_number = int(raw_pr_number)
+        issue_number = int(evidence.get("issue_number"))
+    except (TypeError, ValueError):
+        return None
+
+    return {
+        "issue_number": issue_number,
+        "pr_number": pr_number,
+        "branch_name": branch_name,
+        "head_sha": pr_head,
+    }
+
+
+def build_recovery_record(
+    assessment: Mapping[str, Any],
+    *,
+    recovered_at: str,
+) -> dict[str, Any]:
+    """Durable, secret-free provenance for a completed recovery (#753 AC2/AC6)."""
+    evidence = dict(assessment.get("evidence") or {})
+    return {
+        "recovered": True,
+        "reason": "owning MCP session exited; durable ownership evidence matched",
+        "recovered_at": recovered_at,
+        "prior_session_pid": evidence.get("prior_session_pid"),
+        "replacement_session_pid": evidence.get("replacement_session_pid"),
+        "prior_pid_alive": evidence.get("prior_pid_alive"),
+        "branch_name": evidence.get("locked_branch"),
+        "worktree_path": evidence.get("locked_worktree_path"),
+        "local_head": evidence.get("local_head"),
+        "remote_head": evidence.get("remote_head"),
+        "pr_head": evidence.get("pr_head"),
+        "pr_number": evidence.get("pr_number"),
+        "identity": evidence.get("locked_identity"),
+        "profile": evidence.get("locked_profile"),
+        "proof": list(assessment.get("reasons") or []),
+    }
+
+
+def format_recovery_refusal(assessment: Mapping[str, Any]) -> str:
+    """Single fail-closed message for a refused recovery attempt."""
+    reasons = list(assessment.get("reasons") or []) or [
+        "dead-session lock recovery evidence did not agree"
+    ]
+    return (
+        "Dead-session issue-lock recovery refused: "
+        + "; ".join(reasons)
+        + " (fail closed)"
+    )

issue_lock_worktree.py

@@ -92,8 +92,19 @@ def assess_issue_lock_worktree(
     inspected_git_root: str | None = None,
     base_branch: str | None = None,
     base_branches: frozenset[str] | None = None,
+    recovery_sanctioned: bool = False,
 ) -> dict:
-    """Fail closed when lock preconditions are not met on the declared worktree."""
+    """Fail closed when lock preconditions are not met on the declared worktree.
+
+    ``recovery_sanctioned`` is set only when ``issue_lock_recovery`` has already
+    proven, from the durable lock itself, that this is a dead-session recovery of
+    an existing claim (#753): same issue, branch, worktree, author, and head, with
+    the recording process dead. In that one case the base-equivalence requirement
+    is waived, because a branch that already carries the work is ahead of its base
+    by construction and could never satisfy it. Every other precondition —
+    notably worktree cleanliness — still applies unchanged, and brand-new issue
+    claims keep the full base-equivalence requirement.
+    """
     bases = base_branches or BASE_BRANCHES
     reasons: list[str] = []
     path = (worktree_path or "").strip()
@@ -111,7 +122,11 @@ def assess_issue_lock_worktree(
             f"(dirty files: {', '.join(dirty_files)})"
         )
 
-    if base_equivalent is False:
+    if recovery_sanctioned:
+        # Base-equivalence intentionally not evaluated: ownership was proven
+        # against the durable lock record instead (#753).
+        pass
+    elif base_equivalent is False:
         reasons.append(
             "issue lock worktree must be base-equivalent to one of "
             f"{_base_list(bases)} before implementation work; inspected "
@@ -139,6 +154,7 @@ def assess_issue_lock_worktree(
         inspected_git_root=inspected_git_root,
         base_branch=base_branch,
         base_equivalent=base_equivalent,
+        recovery_sanctioned=recovery_sanctioned,
     )
 
 
@@ -197,6 +213,7 @@ def _assessment(
     inspected_git_root: str | None = None,
     base_branch: str | None = None,
     base_equivalent: bool | None = None,
+    recovery_sanctioned: bool = False,
 ) -> dict:
     return {
         "proven": proven,
@@ -208,6 +225,8 @@ def _assessment(
         "dirty_files": dirty_files,
         "base_branch": base_branch,
         "base_equivalent": base_equivalent,
+        "recovery_sanctioned": recovery_sanctioned,
+        "base_equivalence_waived": bool(recovery_sanctioned),
     }
 
 

issue_work_duplicate_gate.py

@@ -2,7 +2,7 @@
 
 from __future__ import annotations
 
-from typing import Any
+from typing import Any, Mapping
 
 import issue_claim_heartbeat as claim_hb
 
@@ -27,6 +27,116 @@ def _linked_open_pr(issue_number: int, open_prs: list[dict]) -> dict | None:
     return claim_hb._linked_open_pr(issue_number, open_prs)
 
 
+def _pr_links_issue(issue_number: int, pr: Mapping[str, Any]) -> bool:
+    """Same linkage rule ``claim_hb._linked_open_pr`` applies, per PR.
+
+    ``_linked_open_pr`` only yields the *first* match, which cannot answer
+    "is there exactly one linked PR?" — a question the owning-PR exemption
+    below must answer before it can trust any of them.
+    """
+    pattern = _issue_pattern(issue_number)
+    head = (pr.get("head") or {}).get("ref") or ""
+    text = f"{pr.get('title', '')} {pr.get('body', '')}".lower()
+    if pattern in head.lower():
+        return True
+    return (
+        f"closes #{int(issue_number)}" in text
+        or f"fixes #{int(issue_number)}" in text
+    )
+
+
+def _all_linked_open_prs(
+    issue_number: int, open_prs: list[dict]
+) -> list[Mapping[str, Any]]:
+    return [pr for pr in (open_prs or []) if _pr_links_issue(issue_number, pr)]
+
+
+def _assess_owning_pr_exemption(
+    issue_number: int,
+    *,
+    linked_open_prs: list[Mapping[str, Any]],
+    locked_branch: str | None,
+    recovered_owning_pr: Mapping[str, Any] | None,
+) -> tuple[bool, list[str]]:
+    """Is the linked open PR provably the one a sanctioned recovery owns (#755)?
+
+    ``recovered_owning_pr`` is produced by
+    ``issue_lock_recovery.owning_pr_recovery_evidence`` from a completed
+    server-side recovery assessment — it is never a caller-supplied field.
+    Every element is re-checked here against the live PR list this gate was
+    given, so a stale or partial token cannot widen the exemption.
+
+    Returns ``(exempt, diagnostic_reasons)``. Diagnostics are only emitted when
+    a token was offered and rejected, so a blocked caller can see which element
+    of ownership disagreed.
+    """
+    if not recovered_owning_pr:
+        return False, []
+
+    notes: list[str] = []
+    token_issue = recovered_owning_pr.get("issue_number")
+    token_pr = recovered_owning_pr.get("pr_number")
+    token_branch = str(recovered_owning_pr.get("branch_name") or "").strip()
+    token_head = str(recovered_owning_pr.get("head_sha") or "").strip()
+    locked = (locked_branch or "").strip()
+
+    if token_issue is not None and int(token_issue) != int(issue_number):
+        notes.append(
+            f"recovery evidence is for issue #{token_issue}, not "
+            f"#{issue_number} (no owning-PR exemption)"
+        )
+        return False, notes
+    if not locked or not token_branch or locked != token_branch:
+        notes.append(
+            f"recovery evidence branch '{token_branch or 'unknown'}' does not "
+            f"match the branch being locked '{locked or 'unknown'}' "
+            "(no owning-PR exemption)"
+        )
+        return False, notes
+    if len(linked_open_prs) != 1:
+        numbers = ", ".join(
+            f"#{pr.get('number')}" for pr in linked_open_prs
+        ) or "none"
+        notes.append(
+            f"{len(linked_open_prs)} open PRs link issue #{issue_number} "
+            f"({numbers}); recovery may only own exactly one "
+            "(no owning-PR exemption)"
+        )
+        return False, notes
+
+    only = linked_open_prs[0]
+    head_obj = only.get("head") or {}
+    only_number = only.get("number")
+    only_ref = str(head_obj.get("ref") or "").strip()
+    only_sha = str(head_obj.get("sha") or "").strip()
+
+    if token_pr is None or only_number is None or int(only_number) != int(token_pr):
+        notes.append(
+            f"linked open PR #{only_number} is not the recovered owning PR "
+            f"#{token_pr} (no owning-PR exemption)"
+        )
+        return False, notes
+    if only_ref != token_branch:
+        notes.append(
+            f"open PR #{only_number} head branch '{only_ref}' does not match "
+            f"the recovered branch '{token_branch}' (no owning-PR exemption)"
+        )
+        return False, notes
+    if not token_head or not only_sha or only_sha != token_head:
+        notes.append(
+            f"open PR #{only_number} head {only_sha or 'unknown'} does not "
+            f"match the recovered head {token_head or 'unknown'} "
+            "(no owning-PR exemption)"
+        )
+        return False, notes
+
+    return True, [
+        f"open PR #{only_number} is the exact PR already owned by the "
+        f"recovering lock for issue #{issue_number} (branch '{token_branch}', "
+        f"head {token_head}); not duplicate work"
+    ]
+
+
 def _matching_branches(
     issue_number: int,
     branch_names: list[str],
@@ -52,8 +162,15 @@ def assess_work_issue_duplicate_gate(
     claim_entry: dict | None = None,
     locked_branch: str | None = None,
     phase: str = PHASE_LOCK,
+    recovered_owning_pr: Mapping[str, Any] | None = None,
 ) -> dict[str, Any]:
-    """Fail closed when duplicate work is already in flight for an issue."""
+    """Fail closed when duplicate work is already in flight for an issue.
+
+    ``recovered_owning_pr`` (#755) is server-derived evidence that a sanctioned
+    dead-session lock recovery already owns one specific open PR. It exempts
+    *only* that exact PR from the linked-open-PR blocker; every other duplicate
+    signal, and every mismatch, keeps failing closed.
+    """
     reasons: list[str] = []
     outcome = OUTCOME_DUPLICATE_WORK_NOT_PREVENTED
     prs = list(open_prs or [])
@@ -61,12 +178,23 @@ def assess_work_issue_duplicate_gate(
     pattern = _issue_pattern(issue_number)
 
     linked = _linked_open_pr(issue_number, prs)
+    linked_open_prs = _all_linked_open_prs(issue_number, prs)
+    owning_pr_exempted = False
+    exemption_notes: list[str] = []
     if linked:
-        reasons.append(
+        owning_pr_exempted, exemption_notes = _assess_owning_pr_exemption(
-            f"open PR #{linked.get('number')} already covers issue "
+            issue_number,
-            f"#{issue_number} (fail closed)"
+            linked_open_prs=linked_open_prs,
+            locked_branch=locked_branch,
+            recovered_owning_pr=recovered_owning_pr,
         )
-        outcome = OUTCOME_DUPLICATE_PR_PREVENTED
+        if not owning_pr_exempted:
+            reasons.append(
+                f"open PR #{linked.get('number')} already covers issue "
+                f"#{issue_number} (fail closed)"
+            )
+            reasons.extend(exemption_notes)
+            outcome = OUTCOME_DUPLICATE_PR_PREVENTED
 
     conflicting_branches = _matching_branches(
         issue_number, branches, locked_branch=locked_branch
@@ -122,6 +250,9 @@ def assess_work_issue_duplicate_gate(
         "phase": phase,
         "outcome": outcome,
         "linked_open_pr": linked.get("number") if linked else entry.get("linked_open_pr"),
+        "linked_open_pr_count": len(linked_open_prs),
+        "owning_pr_recovery_exempted": owning_pr_exempted,
+        "owning_pr_recovery_notes": list(exemption_notes),
         "conflicting_branches": conflicting_branches,
         "claim_status": status or None,
         "reasons": reasons,

post_merge_moot_lease_gate.py

@@ -0,0 +1,279 @@
+"""Reconciler authorization gate for post-merge moot-lease cleanup (#745).
+
+``gitea_cleanup_post_merge_moot_lease`` (#515) posts a terminal ``phase:
+released`` lease marker — a real, durable mutation of the PR lease ledger.
+Before #745 it was gated on permissions alone (``gitea.read`` to enter,
+``gitea.pr.comment`` to apply) with no canonical task and no role binding, so
+any profile carrying ``gitea.pr.comment`` reached the mutation path while the
+reconciler could not satisfy the operator-required resolve-exact-task ->
+mutation sequence.
+
+This module holds the pure half of that gate:
+
+* the canonical task name and its tool-name alias;
+* an **append-only** in-process ledger of read-only dry-run assessments;
+* ``assess_apply_authorization``, which decides whether an apply may proceed.
+
+Apply is authorized only when all of the following hold:
+
+* the session resolved exactly the cleanup task (no other task substitutes);
+* the active profile role is ``reconciler``;
+* a prior dry run in this session recorded ``lease_moot`` and
+  ``cleanup_allowed`` for the *same* repository, PR, lease session, candidate
+  head and lease marker id;
+* the live assessment still agrees with that evidence, so a lease superseded
+  between the dry run and the apply fails closed;
+* any caller-supplied expectations match the live lease exactly.
+
+Everything else fails closed. The ledger is only ever appended to — a
+superseded dry run stays visible as history instead of being rewritten — which
+keeps the cleanup audit trail append-only end to end.
+"""
+
+from __future__ import annotations
+
+from datetime import datetime, timezone
+from typing import Any
+
+CLEANUP_TASK = "cleanup_post_merge_moot_lease"
+CLEANUP_TOOL_ALIAS = "gitea_cleanup_post_merge_moot_lease"
+REQUIRED_ROLE = "reconciler"
+REQUIRED_PERMISSION = "gitea.pr.comment"
+
+# The read-only assessment stays reachable under gitea.read for every role —
+# the convention shared with cleanup_stale_review_decision_lock and
+# cleanup_obsolete_reviewer_comment_lease — so any namespace can diagnose a
+# stuck lease. Only the apply path demands CLEANUP_TASK + REQUIRED_ROLE.
+ASSESSMENT_PERMISSION = "gitea.read"
+
+_DRY_RUN_LEDGER: list[dict[str, Any]] = []
+
+
+def _norm(value: Any) -> str:
+    return str(value or "").strip()
+
+
+def _norm_comment_id(value: Any) -> int | None:
+    try:
+        return int(value)
+    except (TypeError, ValueError):
+        return None
+
+
+def record_dry_run(
+    *,
+    pr_number: int,
+    repository_slug: str | None,
+    lease_moot: bool,
+    cleanup_allowed: bool,
+    session_id: str | None,
+    candidate_head: str | None,
+    lease_comment_id: Any,
+    recorded_at: datetime | None = None,
+) -> dict[str, Any]:
+    """Append one read-only assessment to the dry-run ledger.
+
+    Never rewrites or removes a prior entry: repeated dry runs accumulate and
+    ``latest_dry_run`` returns the newest matching one.
+    """
+    entry = {
+        "task": CLEANUP_TASK,
+        "pr_number": int(pr_number),
+        "repository_slug": _norm(repository_slug) or None,
+        "lease_moot": bool(lease_moot),
+        "cleanup_allowed": bool(cleanup_allowed),
+        "session_id": _norm(session_id) or None,
+        "candidate_head": _norm(candidate_head) or None,
+        "lease_comment_id": _norm_comment_id(lease_comment_id),
+        "recorded_at": (recorded_at or datetime.now(timezone.utc)).isoformat(),
+    }
+    _DRY_RUN_LEDGER.append(entry)
+    return dict(entry)
+
+
+def dry_run_history() -> tuple[dict[str, Any], ...]:
+    """Immutable view of every recorded dry run, oldest first."""
+    return tuple(dict(entry) for entry in _DRY_RUN_LEDGER)
+
+
+def latest_dry_run(
+    *, pr_number: int, repository_slug: str | None
+) -> dict[str, Any] | None:
+    """Newest dry-run evidence for this repository + PR, or None."""
+    wanted_repo = _norm(repository_slug)
+    for entry in reversed(_DRY_RUN_LEDGER):
+        if entry["pr_number"] != int(pr_number):
+            continue
+        if _norm(entry.get("repository_slug")) != wanted_repo:
+            continue
+        return dict(entry)
+    return None
+
+
+def _reset_for_testing() -> None:
+    """Drop ledger state between tests. Never called by production paths."""
+    _DRY_RUN_LEDGER.clear()
+
+
+def assess_apply_authorization(
+    *,
+    pr_number: int,
+    repository_slug: str | None,
+    resolved_task: str | None,
+    active_role_kind: str | None,
+    assessment: dict[str, Any],
+    evidence: dict[str, Any] | None,
+    expected_session_id: str | None = None,
+    expected_candidate_head: str | None = None,
+    expected_lease_comment_id: Any = None,
+) -> dict[str, Any]:
+    """Decide whether a moot-lease cleanup apply is authorized (fail closed).
+
+    Returns ``{"allowed", "reasons", "blocker_kind", "evidence_matched", ...}``.
+    ``allowed`` is True only when every check passes; each failure contributes a
+    reason so the caller can report all of them together.
+    """
+    reasons: list[str] = []
+    blocker_kind: str | None = None
+
+    def _block(kind: str, reason: str) -> None:
+        nonlocal blocker_kind
+        reasons.append(reason)
+        if blocker_kind is None:
+            blocker_kind = kind
+
+    # 1. Exact resolved cleanup task. Resolving any other task — including a
+    #    sibling reconciler task — does not authorize this mutation.
+    if _norm(resolved_task) != CLEANUP_TASK:
+        _block(
+            "unresolved_cleanup_task",
+            "post-merge moot-lease cleanup requires the session to resolve "
+            f"task '{CLEANUP_TASK}' immediately before apply; resolved task is "
+            f"{resolved_task!r} (fail closed)",
+        )
+
+    # 2. Dedicated reconciler role, enforced independently of the permission.
+    if _norm(active_role_kind) != REQUIRED_ROLE:
+        _block(
+            "wrong_role",
+            f"profile role {active_role_kind!r} cannot apply post-merge "
+            f"moot-lease cleanup; required role is {REQUIRED_ROLE} even when "
+            f"{REQUIRED_PERMISSION} is present (fail closed)",
+        )
+
+    # 3. Canonical repository identity must be established, never inferred from
+    #    request parameters.
+    if not _norm(repository_slug):
+        _block(
+            "repository_binding",
+            "no canonical repository identity could be established for the "
+            "cleanup target (fail closed)",
+        )
+
+    # 4. The live safety assessment must still say the lease is moot/cleanable.
+    if not assessment.get("is_moot") or not assessment.get("cleanup_allowed"):
+        _block(
+            "lease_not_moot",
+            "live assessment does not report a moot, cleanable lease on PR "
+            f"#{pr_number} (lease_moot={bool(assessment.get('is_moot'))}, "
+            f"cleanup_allowed={bool(assessment.get('cleanup_allowed'))}) "
+            "(fail closed)",
+        )
+
+    live = assessment.get("active_lease") or {}
+    live_session = _norm(live.get("session_id"))
+    live_head = _norm(live.get("candidate_head"))
+    live_comment_id = _norm_comment_id(live.get("comment_id"))
+
+    # 5. A lease missing identifying fields is malformed and unsafe to act on.
+    if not live_session or not live_head or live_comment_id is None:
+        _block(
+            "malformed_lease",
+            "active lease is malformed: session_id / candidate_head / "
+            "comment_id must all be present to authorize cleanup "
+            f"(session_id={live.get('session_id')!r}, "
+            f"candidate_head={live.get('candidate_head')!r}, "
+            f"comment_id={live.get('comment_id')!r}) (fail closed)",
+        )
+
+    # 6. Caller expectations, when supplied, must match the live lease exactly.
+    if expected_session_id is not None and _norm(expected_session_id) != live_session:
+        _block(
+            "lease_mismatch",
+            f"expected lease session {expected_session_id!r} does not match the "
+            f"live lease session {live.get('session_id')!r} (fail closed)",
+        )
+    if (
+        expected_candidate_head is not None
+        and _norm(expected_candidate_head) != live_head
+    ):
+        _block(
+            "lease_mismatch",
+            f"expected candidate head {expected_candidate_head!r} does not "
+            f"match the live lease head {live.get('candidate_head')!r} "
+            "(fail closed)",
+        )
+    if expected_lease_comment_id is not None and (
+        _norm_comment_id(expected_lease_comment_id) != live_comment_id
+    ):
+        _block(
+            "lease_mismatch",
+            f"expected lease marker {expected_lease_comment_id!r} does not "
+            f"match the live lease marker {live.get('comment_id')!r} "
+            "(fail closed)",
+        )
+
+    # 7. Matching dry-run evidence recorded earlier in this session.
+    evidence_matched = False
+    if evidence is None:
+        _block(
+            "missing_dry_run_evidence",
+            "no read-only dry run recorded for this repository and PR; run the "
+            "tool with apply=false and confirm lease_moot / cleanup_allowed "
+            "before applying (fail closed)",
+        )
+    elif not evidence.get("lease_moot") or not evidence.get("cleanup_allowed"):
+        _block(
+            "dry_run_not_allowed",
+            "recorded dry run did not report an allowed cleanup "
+            f"(lease_moot={bool(evidence.get('lease_moot'))}, "
+            f"cleanup_allowed={bool(evidence.get('cleanup_allowed'))}) "
+            "(fail closed)",
+        )
+    elif int(evidence.get("pr_number") or -1) != int(pr_number) or _norm(
+        evidence.get("repository_slug")
+    ) != _norm(repository_slug):
+        _block(
+            "dry_run_mismatch",
+            "recorded dry run targets a different repository or PR "
+            f"({evidence.get('repository_slug')}#{evidence.get('pr_number')} vs "
+            f"{repository_slug}#{pr_number}) (fail closed)",
+        )
+    elif (
+        _norm(evidence.get("session_id")) != live_session
+        or _norm(evidence.get("candidate_head")) != live_head
+        or _norm_comment_id(evidence.get("lease_comment_id")) != live_comment_id
+    ):
+        _block(
+            "superseded_lease",
+            "the lease changed after the recorded dry run (dry run: "
+            f"session={evidence.get('session_id')!r}, "
+            f"head={evidence.get('candidate_head')!r}, "
+            f"marker={evidence.get('lease_comment_id')!r}; live: "
+            f"session={live.get('session_id')!r}, "
+            f"head={live.get('candidate_head')!r}, "
+            f"marker={live.get('comment_id')!r}); re-run the dry run "
+            "(fail closed)",
+        )
+    else:
+        evidence_matched = True
+
+    return {
+        "allowed": not reasons,
+        "reasons": reasons,
+        "blocker_kind": blocker_kind,
+        "evidence_matched": evidence_matched,
+        "required_task": CLEANUP_TASK,
+        "required_role_kind": REQUIRED_ROLE,
+        "required_permission": REQUIRED_PERMISSION,
+    }

pr_sync_status.py

@@ -41,6 +41,186 @@ _DENIED_UPDATE_ROLES = frozenset({"reviewer", "merger", "reconciler", "mixed", "
 UPDATE_STYLE_MERGE = "merge"
 _FORBIDDEN_UPDATE_STYLES = frozenset({"rebase", "rebase-merge", "squash", "force"})
 
+# ── Commit check classifications (#751) ──────────────────────────────────
+# Gitea's *combined* commit status reports ``state: pending`` both when a real
+# check is executing and when the status-context collection is empty. Reading
+# ``state`` alone therefore cannot distinguish "CI is running" from "no CI
+# exists", which permanently blocks a merge-ready PR that no check will ever
+# report on. These classifications are derived from the actual context
+# collection plus the live branch-protection policy.
+CHECKS_SUCCESS = "success"
+CHECKS_FAILURE = "failure"
+CHECKS_PENDING = "pending"
+CHECKS_NONE = "none"                          # configured/produced nothing
+CHECKS_NOT_REQUIRED = "not_required"          # protection does not require checks
+CHECKS_MISSING_REQUIRED = "missing_required"  # required contexts have no result
+CHECKS_UNKNOWN = "unknown"                    # indeterminable — fail closed
+
+# Values that permit merge_now when checks are required.
+_CHECKS_OK = frozenset({"success", "passed", "ok", "skipped", "not_required"})
+
+# Raw per-context state vocabularies reported by Gitea.
+_CTX_SUCCESS = frozenset({"success", "passed", "ok"})
+_CTX_FAILURE = frozenset({"failure", "failed", "error", "cancelled", "canceled"})
+_CTX_PENDING = frozenset({"pending", "running", "queued", "expected"})
+_CTX_SKIPPED = frozenset({"skipped", "neutral"})
+
+
+def _normalize_context_rows(statuses: Any) -> list[dict[str, str]]:
+    """Reduce a raw status collection to newest-wins ``{context, state}`` rows.
+
+    Gitea returns the status collection newest-first, so the first row seen for
+    a context wins. Rows without a usable state are discarded rather than being
+    silently treated as passing.
+    """
+    rows: list[dict[str, str]] = []
+    seen: set[str] = set()
+    if not isinstance(statuses, list):
+        return rows
+    for raw in statuses:
+        if not isinstance(raw, dict):
+            continue
+        context = (raw.get("context") or raw.get("name") or "").strip()
+        state = (raw.get("status") or raw.get("state") or "").strip().lower()
+        if not state:
+            continue
+        key = context or f"__unnamed__{len(rows)}"
+        if key in seen:
+            continue
+        seen.add(key)
+        rows.append({"context": context, "state": state})
+    return rows
+
+
+def _aggregate_context_states(rows: list[dict[str, str]]) -> str:
+    """Fail-closed aggregate: failure > pending > unknown-state > success."""
+    states = {row["state"] for row in rows}
+    if states & _CTX_FAILURE:
+        return CHECKS_FAILURE
+    if states & _CTX_PENDING:
+        return CHECKS_PENDING
+    unresolved = states - _CTX_SUCCESS - _CTX_SKIPPED
+    if unresolved:
+        # An unrecognized context state must never read as success.
+        return CHECKS_UNKNOWN
+    return CHECKS_SUCCESS
+
+
+def classify_commit_checks(
+    *,
+    combined_state: str | None = None,
+    statuses: Any = None,
+    checks_enabled: bool | None = None,
+    required_contexts: Any = None,
+    policy_determinable: bool = True,
+    status_determinable: bool = True,
+) -> dict[str, Any]:
+    """Classify head checks from live evidence (#751).
+
+    ``combined_state`` is deliberately **not** authoritative: it is recorded for
+    observability but never used to infer that CI is executing. The context
+    collection and the live protection policy decide.
+
+    Returns ``checks_status`` (one of the ``CHECKS_*`` values), the derived
+    ``checks_required`` flag, and structured ``reasons``.
+    """
+    reasons: list[str] = []
+    rows = _normalize_context_rows(statuses)
+    required = [
+        str(ctx).strip()
+        for ctx in (required_contexts or [])
+        if str(ctx or "").strip()
+    ]
+    observed_combined = (combined_state or "").strip().lower() or None
+
+    result: dict[str, Any] = {
+        "checks_status": CHECKS_UNKNOWN,
+        "checks_required": True,
+        "combined_state": observed_combined,
+        "context_count": len(rows),
+        "observed_contexts": [row["context"] for row in rows],
+        "required_contexts": required,
+        "missing_required_contexts": [],
+        "policy_determinable": bool(policy_determinable),
+        "status_determinable": bool(status_determinable),
+        "reasons": reasons,
+    }
+
+    # Policy unreadable → never assume checks are optional.
+    if not policy_determinable:
+        reasons.append(
+            "branch-protection check policy could not be read; cannot prove "
+            "whether status checks are required (fail closed)"
+        )
+        return result
+
+    if checks_enabled is False:
+        result["checks_required"] = False
+        result["checks_status"] = CHECKS_NOT_REQUIRED
+        reasons.append(
+            "live branch protection does not require status checks for the base "
+            "branch; head status contexts do not gate merge"
+        )
+        return result
+
+    if checks_enabled is None:
+        reasons.append(
+            "branch-protection status-check requirement is indeterminate "
+            "(fail closed)"
+        )
+        return result
+
+    # Checks are required from here on.
+    if not status_determinable:
+        reasons.append(
+            "head commit status collection could not be read while branch "
+            "protection requires status checks (fail closed)"
+        )
+        return result
+
+    if required:
+        by_context = {row["context"]: row["state"] for row in rows if row["context"]}
+        missing = [ctx for ctx in required if ctx not in by_context]
+        if missing:
+            result["missing_required_contexts"] = missing
+            result["checks_status"] = CHECKS_MISSING_REQUIRED
+            reasons.append(
+                "branch protection requires status context(s) "
+                f"{', '.join(missing)} but no matching status result exists at "
+                "the head commit (fail closed)"
+            )
+            return result
+        matched = [
+            {"context": ctx, "state": by_context[ctx]} for ctx in required
+        ]
+        result["checks_status"] = _aggregate_context_states(matched)
+        reasons.append(
+            f"evaluated {len(matched)} required status context(s) from live "
+            "branch protection; unrelated contexts were ignored"
+        )
+        return result
+
+    # Status checks enabled with no specific required contexts configured.
+    if not rows:
+        result["checks_status"] = CHECKS_NONE
+        reasons.append(
+            "branch protection enables status checks but no status context was "
+            "produced for the head commit"
+        )
+        if observed_combined in _CTX_PENDING:
+            reasons.append(
+                f"combined commit state '{observed_combined}' does not indicate "
+                "executing CI because the status-context collection is empty"
+            )
+        return result
+
+    result["checks_status"] = _aggregate_context_states(rows)
+    reasons.append(
+        f"aggregated {len(rows)} reported status context(s); branch protection "
+        "configures no explicit required-context list"
+    )
+    return result
+
 
 def _normalize_sha(value: str | None) -> str | None:
     text = (value or "").strip().lower()
@@ -120,6 +300,7 @@ def assess_pr_sync_status(
         "branch_protection_requires_current_base": requires_current,
         "approval_at_current_head": approval_ok if approval_at_current_head is not None else None,
         "checks_status": checks,
+        "checks_required": bool(checks_required),
         "active_locks_and_leases": {
             "author_lock": bool(active_author_lock) if active_author_lock is not None else None,
             "reviewer_lease": bool(active_reviewer_lease) if active_reviewer_lease is not None else None,
@@ -258,21 +439,41 @@ def assess_pr_sync_status(
         result["recommended_next_action"] = ACTION_BLOCKED
         return result
 
-    # ── Checks gate for merge_now ────────────────────────────────────────
+    # ── Checks gate for merge_now (#751) ─────────────────────────────────
-    if checks_required and checks not in ("success", "passed", "ok", "none", "skipped", "not_required"):
+    # ``checks_required`` is derived from the live branch-protection policy by
-        if checks in ("pending", "running", "queued"):
+    # the production caller. When protection does not require status checks,
+    # head contexts cannot gate the merge and this whole gate is skipped.
+    if not checks_required:
+        reasons.append(
+            "live branch protection does not require status checks; head check "
+            f"state ({checks}) does not gate merge"
+        )
+    elif checks not in _CHECKS_OK:
+        if checks in _CTX_PENDING:
             reasons.append(f"required checks are not finished (status={checks})")
-            result["recommended_next_action"] = ACTION_BLOCKED
+        elif checks in _CTX_FAILURE:
-            return result
-        if checks in ("failure", "failed", "error", "cancelled"):
             reasons.append(f"required checks failed (status={checks})")
-            result["recommended_next_action"] = ACTION_BLOCKED
+        elif checks == CHECKS_MISSING_REQUIRED:
-            return result
+            reasons.append(
-        # unknown — fail closed when checks_required
+                "branch protection configures required status context(s) but no "
-        if checks == "unknown":
+                "matching status result exists at the current head (fail closed)"
+            )
+        elif checks == CHECKS_NONE:
+            reasons.append(
+                "branch protection requires status checks but no status context "
+                "was produced for the current head (fail closed); an empty "
+                "status collection is not executing CI"
+            )
+        elif checks == CHECKS_UNKNOWN:
             reasons.append("checks status unknown (fail closed)")
-            result["recommended_next_action"] = ACTION_BLOCKED
+        else:
-            return result
+            # Unrecognized vocabulary must never fall through to merge_now.
+            reasons.append(
+                f"unrecognized checks status '{checks}' cannot prove required "
+                "checks passed (fail closed)"
+            )
+        result["recommended_next_action"] = ACTION_BLOCKED
+        return result
 
     # ── Ready to merge without update ────────────────────────────────────
     # Includes: current with approval; outdated when update is NOT required.

role_session_router.py

@@ -87,6 +87,11 @@ RECONCILER_TASKS = frozenset({
     # only to the reconciler profile). Raw gitea_delete_branch redirects here to
     # the guarded gitea_cleanup_merged_pr_branch path (#514/#687).
     "delete_branch",
+    # #745: post-merge moot reviewer-lease cleanup is reconciler-owned; the
+    # apply path posts a terminal lease marker. Kept in step with
+    # task_capability_map so map and router cannot disagree (#723 defect A).
+    "cleanup_post_merge_moot_lease",
+    "gitea_cleanup_post_merge_moot_lease",
     "reconcile_already_landed_pr",
     "reconcile_already_landed",
     "reconcile-landed-pr",
@@ -118,6 +123,9 @@ TASK_REQUIRED_ROLE = {
     "reconcile_already_landed": "reconciler",
     "reconcile-landed-pr": "reconciler",
     "cleanup_merged_pr_branch": "reconciler",
+    # #745: post-merge moot reviewer-lease cleanup (canonical task + tool alias).
+    "cleanup_post_merge_moot_lease": "reconciler",
+    "gitea_cleanup_post_merge_moot_lease": "reconciler",
     # #309: reconciler tasks close already-landed PRs/issues only.
     "reconcile_close_landed_pr": "reconciler",
     "reconcile_close_landed_issue": "reconciler",

skills/llm-project-workflow/workflows/create-issue.md

@@ -450,6 +450,35 @@ If any gate fails, do not create the issue.
 
 Produce a recovery handoff or duplicate report.
 
+### 18a. Sanctioned first-mutation path (#749)
+
+`gitea_create_issue` is a **pure remote mutation** (no local tree write). The
+issue-first gate forbids creating `branches/issue-<N>-*` before the issue
+number exists. Therefore the **only sanctioned first mutation** is:
+
+1. Read-only identity + capability + duplicate search from the control checkout.
+2. Ensure the **canonical control checkout** is:
+   * the configured repository root for the requested remote/org/repo;
+   * on an accepted base branch (`master` / `main` / `dev`);
+   * base-equivalent to live master;
+   * clean (no tracked local edits);
+   * in runtime/master parity.
+3. Resolve exact task `create_issue`, then call `gitea_create_issue` **from that
+   clean control checkout** (no `worktree_path` required for this step alone).
+4. After the issue number exists: create a **registered** worktree under
+   `branches/issue-<N>-*`, claim/lock, and perform every subsequent author
+   mutation from that worktree only.
+
+**Forbidden improvisations (fail closed):**
+
+* `mkdir` dummy directories under `branches/` (#713)
+* borrowing an unrelated pre-existing worktree
+* creating a pre-issue worktree in violation of issue-first
+* running create_issue from a dirty, drifted, detached, or non-canonical root
+
+Post-creation mutations (`lock_issue`, commit, push, `create_pr`, etc.) **never**
+receive this bootstrap exemption.
+
 ## 19. Issue commenting gate
 
 Before commenting on an existing issue, verify:

task_capability_map.py

@@ -158,6 +158,23 @@ TASK_CAPABILITY_MAP: dict[str, dict[str, str]] = {
         "permission": "gitea.pr.comment",
         "role": "reviewer",
     },
+    # #745: post-merge moot reviewer-lease cleanup is reconciler-owned. The
+    # apply path posts an append-only terminal `phase: released` lease marker
+    # (gitea.pr.comment), so holding the comment permission alone must not
+    # authorize it — author, reviewer and merger fail closed on the role gate
+    # even though their profiles carry gitea.pr.comment. The read-only
+    # `apply=false` assessment deliberately stays reachable under gitea.read
+    # inside the tool (the same convention as cleanup_stale_review_decision_lock
+    # below), so any namespace can diagnose a stuck lease; only apply requires
+    # this task plus the reconciler role.
+    "cleanup_post_merge_moot_lease": {
+        "permission": "gitea.pr.comment",
+        "role": "reconciler",
+    },
+    "gitea_cleanup_post_merge_moot_lease": {
+        "permission": "gitea.pr.comment",
+        "role": "reconciler",
+    },
     "blind_pr_queue_review": {
         "permission": "gitea.pr.review",
         "role": "reviewer",

tests/test_create_issue_bootstrap.py

@@ -0,0 +1,367 @@
+"""Regression tests for create_issue bootstrap (#749).
+
+TDD: these tests define the sanctioned first-mutation path for
+``gitea_create_issue`` from a clean canonical control checkout, and prove
+the exemption cannot widen to dirty roots, foreign clones, arbitrary
+``branches/`` directories, or post-creation author mutations.
+"""
+
+from __future__ import annotations
+
+import os
+import sys
+import unittest
+from pathlib import Path
+from unittest.mock import MagicMock, patch
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+
+import create_issue_bootstrap as cib  # noqa: E402
+import gitea_mcp_server as srv  # noqa: E402
+import workflow_scope_guard as wsg  # noqa: E402
+
+FAKE_AUTH = {"Authorization": "token test-token"}
+MASTER_SHA = "a" * 40
+STALE_SHA = "b" * 40
+
+current_file_path = Path(__file__).resolve()
+if "branches" in current_file_path.parts:
+    CONTROL_CHECKOUT_ROOT = str(current_file_path.parents[3])
+else:
+    CONTROL_CHECKOUT_ROOT = str(current_file_path.parents[1])
+
+
+class TestCreateIssueBootstrapAssessor(unittest.TestCase):
+    ROOT = "/repo/Gitea-Tools"
+
+    def test_non_create_issue_task_not_applicable(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=self.ROOT,
+            canonical_repo_root=self.ROOT,
+            current_branch="master",
+            head_sha=MASTER_SHA,
+            porcelain_status="",
+            remote_master_sha=MASTER_SHA,
+            task="lock_issue",
+        )
+        self.assertTrue(res["not_applicable"])
+        self.assertFalse(res["allowed"])
+        self.assertFalse(res["block"])
+
+    def test_branches_worktree_not_applicable(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=f"{self.ROOT}/branches/issue-1-x",
+            canonical_repo_root=self.ROOT,
+            current_branch="fix/issue-1-x",
+            task="create_issue",
+        )
+        self.assertTrue(res["not_applicable"])
+        self.assertFalse(res["allowed"])
+
+    def test_clean_control_checkout_allowed(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=self.ROOT,
+            canonical_repo_root=self.ROOT,
+            current_branch="master",
+            head_sha=MASTER_SHA,
+            porcelain_status="",
+            remote_master_sha=MASTER_SHA,
+            task="create_issue",
+        )
+        self.assertFalse(res["not_applicable"])
+        self.assertTrue(res["allowed"])
+        self.assertFalse(res["block"])
+        self.assertEqual(res["bootstrap_path"], "clean_canonical_control_checkout")
+        # Post-create next action must name issue-backed worktree after N exists.
+        self.assertIn("branches/issue-<N>-*", res["exact_next_action"])
+
+    def test_tool_alias_gitea_create_issue_allowed(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=self.ROOT,
+            canonical_repo_root=self.ROOT,
+            current_branch="main",
+            head_sha=MASTER_SHA,
+            porcelain_status="",
+            remote_master_sha=MASTER_SHA,
+            task="gitea_create_issue",
+        )
+        self.assertTrue(res["allowed"])
+
+    def test_dirty_control_checkout_blocked(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=self.ROOT,
+            canonical_repo_root=self.ROOT,
+            current_branch="master",
+            head_sha=MASTER_SHA,
+            porcelain_status=" M gitea_mcp_server.py\n",
+            remote_master_sha=MASTER_SHA,
+            task="create_issue",
+        )
+        self.assertTrue(res["block"])
+        self.assertFalse(res["allowed"])
+        self.assertTrue(any("tracked local edits" in r for r in res["reasons"]))
+        # Pre-issue phase: next action must be satisfiable without inventing <N>.
+        next_a = res["exact_next_action"] or ""
+        self.assertIn("clean accepted base branch", next_a)
+        self.assertIn("before the issue exists", next_a)
+        # Must not prescribe "bind branches/issue-<N>" as the recovery step.
+        self.assertNotIn("Bind an issue-backed worktree", next_a)
+
+    def test_stale_base_blocked(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=self.ROOT,
+            canonical_repo_root=self.ROOT,
+            current_branch="master",
+            head_sha=STALE_SHA,
+            porcelain_status="",
+            remote_master_sha=MASTER_SHA,
+            task="create_issue",
+        )
+        self.assertTrue(res["block"])
+        self.assertTrue(any("live master" in r for r in res["reasons"]))
+
+    def test_non_base_branch_blocked(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=self.ROOT,
+            canonical_repo_root=self.ROOT,
+            current_branch="feat/something",
+            head_sha=MASTER_SHA,
+            porcelain_status="",
+            remote_master_sha=MASTER_SHA,
+            task="create_issue",
+        )
+        self.assertTrue(res["block"])
+
+    def test_detached_head_blocked(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=self.ROOT,
+            canonical_repo_root=self.ROOT,
+            current_branch="",
+            head_sha=MASTER_SHA,
+            porcelain_status="",
+            remote_master_sha=MASTER_SHA,
+            task="create_issue",
+        )
+        self.assertTrue(res["block"])
+        self.assertTrue(any("detached" in r for r in res["reasons"]))
+
+    def test_foreign_workspace_blocked(self):
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path="/other/clone",
+            canonical_repo_root=self.ROOT,
+            current_branch="master",
+            head_sha=MASTER_SHA,
+            porcelain_status="",
+            remote_master_sha=MASTER_SHA,
+            task="create_issue",
+        )
+        self.assertTrue(res["block"])
+        self.assertTrue(any("canonical control checkout" in r for r in res["reasons"]))
+
+
+class TestCreateIssueBootstrapIntegration(unittest.TestCase):
+    def setUp(self):
+        srv._preflight_whoami_called = True
+        srv._preflight_capability_called = True
+        srv._preflight_resolved_role = "author"
+        srv._preflight_resolved_task = "create_issue"
+        srv._preflight_whoami_violation = False
+        srv._preflight_capability_violation = False
+        self._orig_in_test = srv._preflight_in_test_mode
+        srv._preflight_in_test_mode = lambda: False
+
+    def tearDown(self):
+        srv._preflight_in_test_mode = self._orig_in_test
+        srv._preflight_resolved_task = None
+
+    def _git_state(self, branch="master", head=MASTER_SHA, porcelain=""):
+        return {
+            "current_branch": branch,
+            "head_sha": head,
+            "porcelain_status": porcelain,
+        }
+
+    @patch("gitea_mcp_server._auth", return_value=FAKE_AUTH)
+    @patch("gitea_mcp_server._profile_permission_block", return_value=None)
+    @patch("gitea_mcp_server._namespace_mutation_block", return_value=None)
+    @patch(
+        "gitea_mcp_server.role_session_router.check_author_mutation_after_reviewer_stop",
+        return_value=(True, []),
+    )
+    @patch("gitea_mcp_server.api_request")
+    @patch("gitea_mcp_server.api_get_all", return_value=[])
+    @patch(
+        "gitea_mcp_server.root_checkout_guard.resolve_remote_master_sha",
+        return_value=MASTER_SHA,
+    )
+    def test_clean_control_checkout_create_issue_succeeds(
+        self, _remote_sha, _get_all, mock_api, _role, _ns, _prof, _auth
+    ):
+        mock_api.return_value = {
+            "number": 99,
+            "html_url": "https://gitea.example.com/issues/99",
+        }
+        with patch.object(srv, "PROJECT_ROOT", CONTROL_CHECKOUT_ROOT):
+            with patch(
+                "gitea_mcp_server.issue_lock_worktree.read_worktree_git_state",
+                return_value=self._git_state(),
+            ):
+                with patch(
+                    "gitea_mcp_server._get_workspace_porcelain", return_value=""
+                ):
+                    with patch(
+                        "gitea_mcp_server._enforce_root_checkout_guard"
+                    ):
+                        # Anti-stomp / master parity: keep gates green.
+                        with patch.object(
+                            srv,
+                            "_run_anti_stomp_preflight",
+                            return_value=None,
+                        ):
+                            res = srv.gitea_create_issue(
+                                title="Bootstrap issue from clean control",
+                                body="Body text for content gate.",
+                            )
+        self.assertEqual(res.get("number"), 99)
+        mock_api.assert_called_once()
+
+    @patch("gitea_mcp_server._auth", return_value=FAKE_AUTH)
+    @patch("gitea_mcp_server._profile_permission_block", return_value=None)
+    @patch("gitea_mcp_server._namespace_mutation_block", return_value=None)
+    @patch(
+        "gitea_mcp_server.role_session_router.check_author_mutation_after_reviewer_stop",
+        return_value=(True, []),
+    )
+    @patch("gitea_mcp_server.api_request")
+    @patch("gitea_mcp_server.api_get_all", return_value=[])
+    def test_dirty_control_checkout_create_issue_fails_closed(
+        self, _get_all, mock_api, _role, _ns, _prof, _auth
+    ):
+        with patch.object(srv, "PROJECT_ROOT", CONTROL_CHECKOUT_ROOT):
+            with patch(
+                "gitea_mcp_server.issue_lock_worktree.read_worktree_git_state",
+                return_value=self._git_state(
+                    porcelain=" M author_mutation_worktree.py\n"
+                ),
+            ):
+                with patch(
+                    "gitea_mcp_server._get_workspace_porcelain",
+                    return_value=" M author_mutation_worktree.py\n",
+                ):
+                    res = srv.gitea_create_issue(
+                        title="Should fail on dirty root",
+                        body="Body text for content gate.",
+                    )
+        self.assertFalse(res.get("success", True) and res.get("number"))
+        if isinstance(res, dict) and res.get("success") is False:
+            blob = " ".join(res.get("reasons") or [])
+            self.assertTrue(
+                "tracked local edits" in blob
+                or "dirty" in blob.lower()
+                or "control checkout" in blob.lower()
+                or res.get("blocker_kind")
+            )
+            # Pre-issue phase must not demand issue-<N> worktree.
+            next_a = res.get("exact_next_action") or ""
+            if next_a:
+                self.assertNotIn("issue-<N>-*", next_a)
+        mock_api.assert_not_called()
+
+    @patch("gitea_mcp_server._auth", return_value=FAKE_AUTH)
+    @patch("gitea_mcp_server._profile_permission_block", return_value=None)
+    @patch("gitea_mcp_server._namespace_mutation_block", return_value=None)
+    @patch(
+        "gitea_mcp_server.role_session_router.check_author_mutation_after_reviewer_stop",
+        return_value=(True, []),
+    )
+    def test_lock_issue_still_requires_branches_worktree(self, _role, _ns, _prof, _auth):
+        """Existing issue-backed mutations receive no exemption (#749 AC3/AC7)."""
+        srv._preflight_resolved_task = "lock_issue"
+        with patch.object(srv, "PROJECT_ROOT", CONTROL_CHECKOUT_ROOT):
+            with patch(
+                "gitea_mcp_server.issue_lock_worktree.read_worktree_git_state",
+                return_value=self._git_state(),
+            ):
+                with patch(
+                    "gitea_mcp_server._get_workspace_porcelain", return_value=""
+                ):
+                    with patch(
+                        "gitea_mcp_server.root_checkout_guard.resolve_remote_master_sha",
+                        return_value=MASTER_SHA,
+                    ):
+                        with patch.object(
+                            srv, "_enforce_root_checkout_guard"
+                        ):
+                            with self.assertRaises(RuntimeError) as ctx:
+                                srv.verify_preflight_purity(
+                                    remote="prgs",
+                                    task="lock_issue",
+                                )
+        msg = str(ctx.exception)
+        self.assertTrue(
+            "Branches-only mutation guard" in msg
+            or "stable control checkout" in msg,
+            msg,
+        )
+        self.assertIn("control checkout", msg)
+
+    def test_workflow_scope_skips_missing_worktree_for_create_issue_clean_root(self):
+        """#683 root assessor must not block clean-root create_issue bootstrap."""
+        res = wsg.assess_root_source_mutation(
+            workspace_path=CONTROL_CHECKOUT_ROOT,
+            canonical_repo_root=CONTROL_CHECKOUT_ROOT,
+            porcelain_status="",
+            role_kind="author",
+            mutation_task="create_issue",
+        )
+        self.assertFalse(res["block"], res)
+        self.assertTrue(res.get("create_issue_bootstrap") or res["proven"])
+
+    def test_workflow_scope_still_blocks_clean_root_for_lock_issue(self):
+        res = wsg.assess_root_source_mutation(
+            workspace_path=CONTROL_CHECKOUT_ROOT,
+            canonical_repo_root=CONTROL_CHECKOUT_ROOT,
+            porcelain_status="",
+            role_kind="author",
+            mutation_task="lock_issue",
+        )
+        self.assertTrue(res["block"])
+        self.assertEqual(res["blocker_kind"], wsg.BLOCKER_MISSING_WORKTREE)
+
+    def test_arbitrary_branches_directory_not_bootstrap(self):
+        """#713: mkdir fake under branches/ is not the bootstrap path."""
+        fake = os.path.join(CONTROL_CHECKOUT_ROOT, "branches", "fake-mkdir-only")
+        res = cib.assess_create_issue_bootstrap(
+            workspace_path=fake,
+            canonical_repo_root=CONTROL_CHECKOUT_ROOT,
+            current_branch="master",
+            head_sha=MASTER_SHA,
+            porcelain_status="",
+            remote_master_sha=MASTER_SHA,
+            task="create_issue",
+        )
+        # Under branches/ → not bootstrap; ordinary membership/registration applies.
+        self.assertTrue(res["not_applicable"])
+        self.assertFalse(res["allowed"])
+
+
+class TestCreateIssueCapabilityAgreement(unittest.TestCase):
+    def test_map_and_alias_agree_on_create_issue(self):
+        import task_capability_map as tcm
+
+        self.assertEqual(
+            tcm.required_permission("create_issue"),
+            "gitea.issue.create",
+        )
+        # Tool alias must resolve to the same task contract.
+        alias = getattr(tcm, "TOOL_TASK_ALIASES", None) or getattr(
+            tcm, "TASK_ALIASES", None
+        )
+        if alias is not None:
+            mapped = alias.get("gitea_create_issue")
+            if mapped is not None:
+                self.assertEqual(mapped, "create_issue")
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_create_issue_workspace_guard.py

@@ -51,29 +51,62 @@ class TestCreateIssueWorkspaceGuard(unittest.TestCase):
             "porcelain_status": "",
         },
     )
-    def test_create_issue_stable_checkout_rejected(
+    def test_create_issue_stable_checkout_bootstrap_allowed_when_clean(
         self, _git, _remote_sha, _get_all, mock_api, _role, _ns, _prof, _auth,
     ):
-        # Without worktree_path/env hints, workspace resolves to PROJECT_ROOT. When that
+        # #749: clean canonical control checkout is the sanctioned create_issue path.
-        # path is the stable control checkout (not under branches/), mutation must fail.
+        mock_api.return_value = {
+            "number": 77,
+            "html_url": "https://gitea.example.com/issues/77",
+        }
         with patch.object(srv, "PROJECT_ROOT", CONTROL_CHECKOUT_ROOT):
-            try:
+            with patch("gitea_mcp_server._get_workspace_porcelain", return_value=""):
-                res = srv.gitea_create_issue(title="Test issue", body="body text")
+                with patch.object(srv, "_run_anti_stomp_preflight", return_value=None):
-            except RuntimeError as exc:
+                    with patch.object(srv, "_enforce_root_checkout_guard"):
-                self.assertIn("stable control checkout", str(exc))
+                        res = srv.gitea_create_issue(
-            else:
+                            title="Test issue", body="body text for gate"
-                # #683: production guards return typed blockers at entrypoints
+                        )
-                self.assertFalse(res.get("success"))
+        self.assertEqual(res.get("number"), 77)
-                self.assertFalse(res.get("performed"))
+        mock_api.assert_called_once()
-                blob = " ".join(res.get("reasons") or []) + " " + str(
+
-                    res.get("blocker_kind") or ""
+    @patch("gitea_mcp_server._auth", return_value=FAKE_AUTH)
-                )
+    @patch("gitea_mcp_server._profile_permission_block", return_value=None)
-                self.assertTrue(
+    @patch("gitea_mcp_server._namespace_mutation_block", return_value=None)
-                    "stable control checkout" in blob
+    @patch("gitea_mcp_server.role_session_router.check_author_mutation_after_reviewer_stop", return_value=(True, []))
-                    or "missing_issue_worktree" in blob
+    @patch("gitea_mcp_server.api_request")
-                    or "control checkout" in blob.lower()
+    @patch("gitea_mcp_server.api_get_all", return_value=[])
-                )
+    @patch("gitea_mcp_server.root_checkout_guard.resolve_remote_master_sha", return_value="a" * 40)
-                self.assertTrue(res.get("exact_next_action"))
+    @patch(
+        "gitea_mcp_server.issue_lock_worktree.read_worktree_git_state",
+        return_value={
+            "current_branch": "master",
+            "head_sha": "a" * 40,
+            "porcelain_status": " M dirty.py\n",
+        },
+    )
+    def test_create_issue_dirty_control_checkout_rejected(
+        self, _git, _remote_sha, _get_all, mock_api, _role, _ns, _prof, _auth,
+    ):
+        # #749: dirty control checkout still fails closed (no bootstrap).
+        with patch.object(srv, "PROJECT_ROOT", CONTROL_CHECKOUT_ROOT):
+            with patch(
+                "gitea_mcp_server._get_workspace_porcelain",
+                return_value=" M dirty.py\n",
+            ):
+                try:
+                    res = srv.gitea_create_issue(title="Test issue", body="body text")
+                except RuntimeError as exc:
+                    self.assertTrue(
+                        "tracked local edits" in str(exc)
+                        or "dirty" in str(exc).lower()
+                        or "bootstrap" in str(exc).lower()
+                        or "control checkout" in str(exc).lower()
+                    )
+                else:
+                    self.assertFalse(res.get("success", True) and res.get("number"))
+                    blob = " ".join(res.get("reasons") or [])
+                    self.assertTrue(blob or res.get("blocker_kind"))
+        mock_api.assert_not_called()
 
     @patch("gitea_mcp_server._auth", return_value=FAKE_AUTH)
     @patch("gitea_mcp_server._profile_permission_block", return_value=None)

tests/test_issue_745_moot_lease_reconciler_gate.py

@@ -0,0 +1,658 @@
+"""Reconciler role binding for post-merge moot-lease cleanup (#745).
+
+``gitea_cleanup_post_merge_moot_lease`` (#515) posts a terminal ``phase:
+released`` lease marker but had no capability-map entry and no role gate: entry
+required only ``gitea.read``, apply required only ``gitea.pr.comment``, so any
+profile holding the comment permission reached the mutation while the
+reconciler could not satisfy resolve-exact-task -> mutation.
+
+These tests pin the fixed contract:
+
+- the canonical task and its tool-name alias resolve identically
+  (``gitea.pr.comment`` + ``reconciler``);
+- only a reconciler profile satisfies permission AND role;
+- ``apply=false`` assessment stays reachable under ``gitea.read`` for any role
+  and mutates nothing (documented, deliberate divergence from the apply path);
+- ``apply=true`` requires the exact resolved task, the reconciler role, the
+  comment permission, a validated repository binding and matching dry-run
+  evidence;
+- live/non-moot/superseded/mismatched/malformed leases fail closed;
+- the dry-run ledger is append-only and cleanup stays idempotent.
+
+Every fixture is synthetic. No production PR, lease session or marker is used
+anywhere in this module (see ``TestNoProductionLeaseTouched``).
+"""
+
+import sys as _sys
+from pathlib import Path as _Path
+
+_sys.path.insert(0, str(_Path(__file__).resolve().parent.parent))
+
+import os  # noqa: E402
+import unittest  # noqa: E402
+from datetime import datetime, timezone  # noqa: E402
+from unittest.mock import patch  # noqa: E402
+
+import mcp_server  # noqa: E402
+import post_merge_moot_lease_gate as gate  # noqa: E402
+import reviewer_pr_lease as leases  # noqa: E402
+from mcp_server import gitea_cleanup_post_merge_moot_lease  # noqa: E402
+from role_session_router import RECONCILER_TASKS, TASK_REQUIRED_ROLE  # noqa: E402
+from task_capability_map import required_permission, required_role  # noqa: E402
+
+FAKE_AUTH = "Basic dGVzdDp0ZXN0"
+SLUG = "Scaled-Tech-Consulting/Gitea-Tools"
+PR = 487
+ISSUE = 485
+SESSION = "97274-676d20a825c4"
+HEAD_A = "a" * 40
+HEAD_B = "d" * 40
+LEASE_COMMENT_ID = 6603
+
+CANONICAL_TASK = "cleanup_post_merge_moot_lease"
+TOOL_ALIAS = "gitea_cleanup_post_merge_moot_lease"
+
+_BASE_OPS = "gitea.read,gitea.pr.comment"
+RECONCILER_ENV = {
+    "GITEA_PROFILE_NAME": "prgs-reconciler",
+    "GITEA_ALLOWED_OPERATIONS": _BASE_OPS + ",gitea.pr.close,gitea.branch.delete",
+}
+AUTHOR_ENV = {
+    "GITEA_PROFILE_NAME": "prgs-author",
+    "GITEA_ALLOWED_OPERATIONS": _BASE_OPS + ",gitea.pr.create,gitea.issue.create",
+}
+REVIEWER_ENV = {
+    "GITEA_PROFILE_NAME": "prgs-reviewer",
+    "GITEA_ALLOWED_OPERATIONS": _BASE_OPS + ",gitea.pr.review,gitea.pr.approve",
+}
+MERGER_ENV = {
+    "GITEA_PROFILE_NAME": "prgs-merger",
+    "GITEA_ALLOWED_OPERATIONS": _BASE_OPS + ",gitea.pr.merge",
+}
+
+# Permission shape of the configured role profiles (mirrors
+# tests/test_task_capability_role_invariants.py CANONICAL_ROLE_PROFILES).
+ROLE_PROFILE_PERMISSIONS = {
+    "author": {"gitea.read", "gitea.pr.comment", "gitea.pr.create",
+               "gitea.issue.create", "gitea.issue.comment", "gitea.issue.close",
+               "gitea.branch.create", "gitea.branch.push", "gitea.repo.commit"},
+    "reviewer": {"gitea.read", "gitea.pr.comment", "gitea.pr.review",
+                 "gitea.pr.approve", "gitea.pr.request_changes",
+                 "gitea.issue.comment"},
+    "merger": {"gitea.read", "gitea.pr.comment", "gitea.pr.merge",
+               "gitea.issue.comment"},
+    "reconciler": {"gitea.read", "gitea.pr.comment", "gitea.pr.close",
+                   "gitea.issue.close", "gitea.branch.delete"},
+}
+
+
+def _lease_comment(pr_number=PR, session_id=SESSION, *, phase="claimed",
+                   candidate_head=HEAD_A, comment_id=LEASE_COMMENT_ID):
+    body = leases.format_lease_body(
+        repo=SLUG,
+        pr_number=pr_number,
+        issue_number=ISSUE,
+        reviewer_identity="sysadmin",
+        profile="prgs-reviewer",
+        session_id=session_id,
+        worktree="branches/review-pr487",
+        phase=phase,
+        candidate_head=candidate_head,
+        target_branch="master",
+        target_branch_sha="b" * 40,
+        last_activity=datetime.now(timezone.utc),
+    )
+    return {"id": comment_id, "body": body, "user": {"login": "sysadmin"}}
+
+
+def _api_side_effect(*, pr_state, pr_merged, comments, posted_id=9999):
+    """api_request side effect keyed on method + url; records POSTs."""
+    calls = {"post": []}
+
+    def _side(method, url, auth=None, payload=None, *a, **k):
+        if (method or "").upper() == "POST":
+            calls["post"].append({"url": url, "payload": payload})
+            return {"id": posted_id}
+        if "/comments" in url:
+            return list(comments)
+        if "/pulls/" in url:
+            pr = {"state": pr_state, "number": PR, "merge_commit_sha": "c" * 40}
+            if pr_merged:
+                pr["merged"] = True
+                pr["merged_at"] = "2026-07-08T07:46:04Z"
+            return pr
+        if "/issues/" in url:
+            return {"state": "closed" if pr_merged else "open", "number": ISSUE}
+        return {}
+
+    return _side, calls
+
+
+def _assessment(comments, *, pr_merged=True, pr_state="closed"):
+    return leases.assess_post_merge_moot_lease(
+        comments, pr_number=PR, pr_merged=pr_merged, pr_state=pr_state,
+        merge_commit_sha="c" * 40,
+    )
+
+
+# --------------------------------------------------------------------------- #
+# 1-5. Capability map / router contract
+# --------------------------------------------------------------------------- #
+class TestCleanupTaskContract(unittest.TestCase):
+    def test_canonical_task_is_reconciler_owned(self):
+        """1. The reconciler is the role that can resolve the cleanup task."""
+        self.assertEqual(required_permission(CANONICAL_TASK), "gitea.pr.comment")
+        self.assertEqual(required_role(CANONICAL_TASK), "reconciler")
+
+    def test_tool_alias_resolves_to_identical_contract(self):
+        """5. Alias and canonical task must not diverge."""
+        self.assertEqual(
+            (required_permission(TOOL_ALIAS), required_role(TOOL_ALIAS)),
+            (required_permission(CANONICAL_TASK), required_role(CANONICAL_TASK)),
+        )
+
+    def test_author_reviewer_merger_cannot_resolve_the_task(self):
+        """2-4. No non-reconciler role satisfies permission AND role."""
+        for role in ("author", "reviewer", "merger"):
+            with self.subTest(role=role):
+                self.assertNotEqual(required_role(CANONICAL_TASK), role)
+                # They hold the permission — which is exactly why the role gate
+                # is required rather than optional.
+                self.assertIn(
+                    "gitea.pr.comment", ROLE_PROFILE_PERMISSIONS[role],
+                    "test premise: non-reconciler roles do hold pr.comment",
+                )
+
+    def test_reconciler_profile_satisfies_permission_and_role(self):
+        self.assertIn(
+            required_permission(CANONICAL_TASK),
+            ROLE_PROFILE_PERMISSIONS[required_role(CANONICAL_TASK)],
+        )
+
+    def test_router_agrees_with_capability_map(self):
+        for task in (CANONICAL_TASK, TOOL_ALIAS):
+            with self.subTest(task=task):
+                self.assertIn(task, RECONCILER_TASKS)
+                self.assertEqual(TASK_REQUIRED_ROLE[task], required_role(task))
+
+    def test_unknown_alias_still_rejected(self):
+        for bogus in ("cleanup_post_merge_moot_leases", "cleanup_moot_lease", ""):
+            with self.subTest(task=bogus):
+                with self.assertRaises(KeyError):
+                    required_role(bogus)
+                with self.assertRaises(KeyError):
+                    required_permission(bogus)
+
+
+# --------------------------------------------------------------------------- #
+# Authorization gate unit tests
+# --------------------------------------------------------------------------- #
+class TestApplyAuthorizationGate(unittest.TestCase):
+    def setUp(self):
+        gate._reset_for_testing()
+        self.addCleanup(gate._reset_for_testing)
+        self.assessment = _assessment([_lease_comment()])
+
+    def _evidence(self, **over):
+        base = dict(
+            pr_number=PR, repository_slug=SLUG, lease_moot=True,
+            cleanup_allowed=True, session_id=SESSION, candidate_head=HEAD_A,
+            lease_comment_id=LEASE_COMMENT_ID,
+        )
+        base.update(over)
+        return gate.record_dry_run(**base)
+
+    def _assess(self, **over):
+        kwargs = dict(
+            pr_number=PR, repository_slug=SLUG, resolved_task=CANONICAL_TASK,
+            active_role_kind="reconciler", assessment=self.assessment,
+            evidence=self._evidence(),
+        )
+        kwargs.update(over)
+        return gate.assess_apply_authorization(**kwargs)
+
+    def test_reconciler_with_matching_evidence_is_authorized(self):
+        result = self._assess()
+        self.assertTrue(result["allowed"], result["reasons"])
+        self.assertTrue(result["evidence_matched"])
+
+    def test_apply_without_exact_task_resolution_fails(self):
+        """8. Apply without exact task resolution fails preflight."""
+        result = self._assess(resolved_task=None)
+        self.assertFalse(result["allowed"])
+        self.assertEqual(result["blocker_kind"], "unresolved_cleanup_task")
+
+    def test_resolving_another_task_does_not_authorize_cleanup(self):
+        """9. A sibling reconciler task is not a substitute."""
+        for other in ("delete_branch", "reconcile_already_landed_pr",
+                      "cleanup_merged_pr_branch", "comment_pr"):
+            with self.subTest(task=other):
+                result = self._assess(resolved_task=other)
+                self.assertFalse(result["allowed"])
+                self.assertEqual(
+                    result["blocker_kind"], "unresolved_cleanup_task")
+
+    def test_non_reconciler_roles_fail_closed(self):
+        """10. Author/reviewer/merger cannot apply despite pr.comment."""
+        for role in ("author", "reviewer", "merger", None, ""):
+            with self.subTest(role=role):
+                result = self._assess(active_role_kind=role)
+                self.assertFalse(result["allowed"])
+                self.assertEqual(result["blocker_kind"], "wrong_role")
+
+    def test_missing_repository_identity_fails_closed(self):
+        result = self._assess(repository_slug=None)
+        self.assertFalse(result["allowed"])
+        self.assertEqual(result["blocker_kind"], "repository_binding")
+
+    def test_missing_dry_run_evidence_fails_closed(self):
+        result = self._assess(evidence=None)
+        self.assertFalse(result["allowed"])
+        self.assertEqual(result["blocker_kind"], "missing_dry_run_evidence")
+
+    def test_dry_run_that_disallowed_cleanup_fails_closed(self):
+        result = self._assess(evidence=self._evidence(cleanup_allowed=False))
+        self.assertFalse(result["allowed"])
+        self.assertEqual(result["blocker_kind"], "dry_run_not_allowed")
+
+    def test_dry_run_for_another_pr_or_repo_fails_closed(self):
+        """11. Wrong PR / repository fails closed."""
+        for over in ({"pr_number": PR + 1}, {"repository_slug": "Other/Repo"}):
+            with self.subTest(**over):
+                result = self._assess(evidence=self._evidence(**over))
+                self.assertFalse(result["allowed"])
+                self.assertEqual(result["blocker_kind"], "dry_run_mismatch")
+
+    def test_superseded_lease_fails_closed(self):
+        """12. Head/session/marker drift since the dry run fails closed."""
+        for over in ({"session_id": "other-session"},
+                     {"candidate_head": HEAD_B},
+                     {"lease_comment_id": 7777}):
+            with self.subTest(**over):
+                result = self._assess(evidence=self._evidence(**over))
+                self.assertFalse(result["allowed"])
+                self.assertEqual(result["blocker_kind"], "superseded_lease")
+
+    def test_expectation_mismatch_fails_closed(self):
+        """11. Wrong session / head / marker expectations fail closed."""
+        for over in ({"expected_session_id": "nope"},
+                     {"expected_candidate_head": HEAD_B},
+                     {"expected_lease_comment_id": 7777}):
+            with self.subTest(**over):
+                result = self._assess(**over)
+                self.assertFalse(result["allowed"])
+                self.assertEqual(result["blocker_kind"], "lease_mismatch")
+
+    def test_matching_expectations_are_authorized(self):
+        result = self._assess(
+            expected_session_id=SESSION,
+            expected_candidate_head=HEAD_A,
+            expected_lease_comment_id=LEASE_COMMENT_ID,
+        )
+        self.assertTrue(result["allowed"], result["reasons"])
+
+    def test_non_moot_lease_fails_closed(self):
+        """12. A live lease on an open PR is never cleanable."""
+        open_pr = _assessment(
+            [_lease_comment()], pr_merged=False, pr_state="open")
+        result = self._assess(assessment=open_pr)
+        self.assertFalse(result["allowed"])
+        self.assertIn(
+            result["blocker_kind"], ("lease_not_moot", "malformed_lease"))
+
+    def test_malformed_lease_fails_closed(self):
+        malformed = dict(self.assessment)
+        malformed["active_lease"] = {
+            "session_id": "", "candidate_head": None, "comment_id": None}
+        result = self._assess(assessment=malformed)
+        self.assertFalse(result["allowed"])
+        self.assertEqual(result["blocker_kind"], "malformed_lease")
+
+    def test_ledger_is_append_only(self):
+        """14. Recording never rewrites or drops prior entries."""
+        first = self._evidence()
+        second = self._evidence(candidate_head=HEAD_B, lease_comment_id=7777)
+        history = gate.dry_run_history()
+        self.assertEqual(len(history), 2)
+        self.assertEqual(history[0]["candidate_head"], first["candidate_head"])
+        self.assertEqual(history[1]["candidate_head"], HEAD_B)
+        # Newest-wins for lookup, but the older entry survives in history.
+        latest = gate.latest_dry_run(pr_number=PR, repository_slug=SLUG)
+        self.assertEqual(latest["lease_comment_id"], second["lease_comment_id"])
+        self.assertEqual(gate.dry_run_history()[0]["lease_comment_id"],
+                         LEASE_COMMENT_ID)
+
+    def test_history_view_cannot_mutate_the_ledger(self):
+        self._evidence()
+        snapshot = gate.dry_run_history()
+        snapshot[0]["pr_number"] = 999999
+        self.assertEqual(
+            gate.dry_run_history()[0]["pr_number"], PR,
+            "dry_run_history must hand out copies, not live rows")
+
+
+# --------------------------------------------------------------------------- #
+# Tool-level behavior
+# --------------------------------------------------------------------------- #
+class _ToolCase(unittest.TestCase):
+    def setUp(self):
+        leases.clear_session_lease()
+        gate._reset_for_testing()
+        self.addCleanup(gate._reset_for_testing)
+
+    def _run(self, env, *, apply, comments, pr_state="closed", pr_merged=True,
+             resolved_task=CANONICAL_TASK, slug=SLUG, **kwargs):
+        side, calls = _api_side_effect(
+            pr_state=pr_state, pr_merged=pr_merged, comments=comments)
+        with patch("mcp_server.api_request", side_effect=side), \
+                patch("mcp_server.get_auth_header", return_value=FAKE_AUTH), \
+                patch("mcp_server.verify_preflight_purity", return_value=None), \
+                patch("mcp_server._bound_repository_slug", return_value=slug), \
+                patch("mcp_server._repository_binding_block", return_value=None), \
+                patch.object(mcp_server, "_preflight_resolved_task",
+                             resolved_task), \
+                patch.dict(os.environ, env, clear=True):
+            result = gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=apply, remote="prgs", **kwargs)
+        return result, calls
+
+
+class TestDryRunOpenToEveryRole(_ToolCase):
+    """7. Dry run performs no mutation and stays under the read capability."""
+
+    def test_dry_run_reports_moot_and_mutates_nothing_for_every_role(self):
+        for name, env in (("reconciler", RECONCILER_ENV), ("author", AUTHOR_ENV),
+                          ("reviewer", REVIEWER_ENV), ("merger", MERGER_ENV)):
+            with self.subTest(role=name):
+                gate._reset_for_testing()
+                result, calls = self._run(
+                    env, apply=False, comments=[_lease_comment()],
+                    resolved_task=None)
+                self.assertTrue(result["success"])
+                self.assertTrue(result["lease_moot"])
+                self.assertTrue(result["cleanup_allowed"])
+                self.assertFalse(result["cleanup_performed"])
+                self.assertEqual(result["mode"], "read_only")
+                self.assertEqual(calls["post"], [], "dry run must not mutate")
+
+    def test_dry_run_records_evidence(self):
+        result, _ = self._run(
+            RECONCILER_ENV, apply=False, comments=[_lease_comment()])
+        evidence = result["dry_run_evidence"]
+        self.assertEqual(evidence["pr_number"], PR)
+        self.assertEqual(evidence["repository_slug"], SLUG)
+        self.assertEqual(evidence["session_id"], SESSION)
+        self.assertEqual(evidence["candidate_head"], HEAD_A)
+        self.assertEqual(evidence["lease_comment_id"], LEASE_COMMENT_ID)
+        self.assertTrue(evidence["lease_moot"])
+        self.assertTrue(evidence["cleanup_allowed"])
+
+
+class TestApplyRequiresReconciler(_ToolCase):
+    def test_reconciler_apply_succeeds_after_matching_dry_run(self):
+        """7 (apply). Allowed dry run then apply posts exactly one marker."""
+        comments = [_lease_comment()]
+        dry, dry_calls = self._run(
+            RECONCILER_ENV, apply=False, comments=comments)
+        self.assertTrue(dry["cleanup_allowed"])
+        self.assertEqual(dry_calls["post"], [])
+
+        result, calls = self._run(
+            RECONCILER_ENV, apply=True, comments=comments)
+        self.assertTrue(result["success"], result.get("reasons"))
+        self.assertTrue(result["cleanup_performed"])
+        self.assertEqual(result["released_comment_id"], 9999)
+        self.assertEqual(len(calls["post"]), 1)
+        body = calls["post"][0]["payload"]["body"]
+        self.assertIn("phase: released", body)
+        self.assertIn("post-merge-moot", body)
+
+    def test_apply_without_dry_run_fails_closed(self):
+        """6. Apply must be preceded by a matching dry run."""
+        result, calls = self._run(
+            RECONCILER_ENV, apply=True, comments=[_lease_comment()])
+        self.assertFalse(result["success"])
+        self.assertFalse(result["cleanup_performed"])
+        self.assertEqual(result["blocker_kind"], "missing_dry_run_evidence")
+        self.assertEqual(calls["post"], [])
+
+    def test_apply_without_exact_task_resolution_fails_closed(self):
+        """8. No resolved cleanup task -> no mutation."""
+        comments = [_lease_comment()]
+        self._run(RECONCILER_ENV, apply=False, comments=comments)
+        result, calls = self._run(
+            RECONCILER_ENV, apply=True, comments=comments, resolved_task=None)
+        self.assertFalse(result["success"])
+        self.assertEqual(result["blocker_kind"], "unresolved_cleanup_task")
+        self.assertEqual(calls["post"], [])
+
+    def test_resolving_a_different_task_does_not_authorize_apply(self):
+        """9. Another resolved task is not a substitute."""
+        comments = [_lease_comment()]
+        self._run(RECONCILER_ENV, apply=False, comments=comments)
+        result, calls = self._run(
+            RECONCILER_ENV, apply=True, comments=comments,
+            resolved_task="delete_branch")
+        self.assertFalse(result["success"])
+        self.assertEqual(result["blocker_kind"], "unresolved_cleanup_task")
+        self.assertEqual(calls["post"], [])
+
+    def test_author_reviewer_merger_cannot_apply(self):
+        """10. Permission-only roles are refused at the role gate."""
+        for name, env in (("author", AUTHOR_ENV), ("reviewer", REVIEWER_ENV),
+                          ("merger", MERGER_ENV)):
+            with self.subTest(role=name):
+                gate._reset_for_testing()
+                comments = [_lease_comment()]
+                self._run(env, apply=False, comments=comments)
+                result, calls = self._run(env, apply=True, comments=comments)
+                self.assertFalse(result["success"])
+                self.assertFalse(result["cleanup_performed"])
+                self.assertEqual(result["blocker_kind"], "wrong_role")
+                self.assertEqual(
+                    calls["post"], [],
+                    f"{name} must not post a terminal lease marker")
+
+
+class TestApplyFailsClosedOnLeaseState(_ToolCase):
+    def test_open_pr_lease_is_never_force_cleaned(self):
+        """12. Non-moot: an active lease on an open PR stays untouched."""
+        comments = [_lease_comment()]
+        result, calls = self._run(
+            RECONCILER_ENV, apply=True, comments=comments,
+            pr_state="open", pr_merged=False)
+        self.assertFalse(result["cleanup_performed"])
+        self.assertFalse(result["pr_merged_or_closed"])
+        self.assertEqual(calls["post"], [], "never force-clean an open PR lease")
+        self.assertTrue(any(
+            "still open" in r for r in result.get("cleanup_skipped_reason", [])))
+
+    def test_superseded_lease_between_dry_run_and_apply_fails_closed(self):
+        """12. The lease moved on after the dry run -> refuse."""
+        self._run(RECONCILER_ENV, apply=False, comments=[_lease_comment()])
+        moved = [_lease_comment(session_id="fresh-session",
+                                candidate_head=HEAD_B, comment_id=7777)]
+        result, calls = self._run(RECONCILER_ENV, apply=True, comments=moved)
+        self.assertFalse(result["success"])
+        self.assertEqual(result["blocker_kind"], "superseded_lease")
+        self.assertEqual(calls["post"], [])
+
+    def test_expectation_mismatch_fails_closed(self):
+        """11. Wrong session / head / marker expectations refuse the apply."""
+        comments = [_lease_comment()]
+        for kwargs in ({"expected_session_id": "wrong-session"},
+                       {"expected_candidate_head": HEAD_B},
+                       {"expected_lease_comment_id": 7777}):
+            with self.subTest(**kwargs):
+                gate._reset_for_testing()
+                self._run(RECONCILER_ENV, apply=False, comments=comments)
+                result, calls = self._run(
+                    RECONCILER_ENV, apply=True, comments=comments, **kwargs)
+                self.assertFalse(result["success"])
+                self.assertEqual(result["blocker_kind"], "lease_mismatch")
+                self.assertEqual(calls["post"], [])
+
+    def test_already_terminal_cleanup_is_idempotent(self):
+        """13. A released lease reports nothing to clean and posts nothing."""
+        first = _assessment([_lease_comment()])
+        released = {"id": 7000, "body": first["release_body"],
+                    "user": {"login": "sysadmin"}}
+        comments = [_lease_comment(), released]
+        self._run(RECONCILER_ENV, apply=False, comments=comments)
+        result, calls = self._run(RECONCILER_ENV, apply=True, comments=comments)
+        self.assertFalse(result["cleanup_performed"])
+        self.assertFalse(result["lease_moot"])
+        self.assertEqual(calls["post"], [], "no second terminal marker")
+        self.assertTrue(any(
+            "already released/terminal" in r
+            for r in result.get("cleanup_skipped_reason", [])))
+
+    def test_apply_is_append_only_never_deletes(self):
+        """14. The only write is a POST; nothing is edited or deleted."""
+        comments = [_lease_comment()]
+        self._run(RECONCILER_ENV, apply=False, comments=comments)
+        side, _calls = _api_side_effect(
+            pr_state="closed", pr_merged=True, comments=comments)
+        seen = []
+
+        def _recording(method, url, auth=None, payload=None, *a, **k):
+            seen.append((method or "").upper())
+            return side(method, url, auth, payload, *a, **k)
+
+        with patch("mcp_server.api_request", side_effect=_recording), \
+                patch("mcp_server.get_auth_header", return_value=FAKE_AUTH), \
+                patch("mcp_server.verify_preflight_purity", return_value=None), \
+                patch("mcp_server._bound_repository_slug", return_value=SLUG), \
+                patch("mcp_server._repository_binding_block", return_value=None), \
+                patch.object(mcp_server, "_preflight_resolved_task",
+                             CANONICAL_TASK), \
+                patch.dict(os.environ, RECONCILER_ENV, clear=True):
+            result = gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=True, remote="prgs")
+        self.assertTrue(result["cleanup_performed"])
+        self.assertNotIn("DELETE", seen)
+        self.assertNotIn("PATCH", seen)
+        self.assertNotIn("PUT", seen)
+        self.assertEqual(seen.count("POST"), 1)
+
+
+class TestRepositoryBinding(_ToolCase):
+    """11. Foreign-repository targets fail closed before any mutation."""
+
+    def test_explicit_foreign_repository_is_rejected(self):
+        comments = [_lease_comment()]
+        side, calls = _api_side_effect(
+            pr_state="closed", pr_merged=True, comments=comments)
+        with patch("mcp_server.api_request", side_effect=side), \
+                patch("mcp_server.get_auth_header", return_value=FAKE_AUTH), \
+                patch("mcp_server.verify_preflight_purity", return_value=None), \
+                patch("mcp_server._canonical_repository_slug",
+                      return_value=(None, [])), \
+                patch("mcp_server._workspace_repository_slug",
+                      return_value=SLUG), \
+                patch.object(mcp_server, "_preflight_resolved_task",
+                             CANONICAL_TASK), \
+                patch.dict(os.environ, RECONCILER_ENV, clear=True):
+            gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=False, remote="prgs")
+            result = gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=True, remote="prgs",
+                org="Some-Other-Org", repo="Some-Other-Repo")
+        self.assertFalse(result["success"])
+        self.assertEqual(result["blocker_kind"], "repository_binding")
+        self.assertEqual(calls["post"], [])
+
+    def test_unresolvable_canonical_root_fails_closed(self):
+        comments = [_lease_comment()]
+        side, calls = _api_side_effect(
+            pr_state="closed", pr_merged=True, comments=comments)
+        with patch("mcp_server.api_request", side_effect=side), \
+                patch("mcp_server.get_auth_header", return_value=FAKE_AUTH), \
+                patch("mcp_server.verify_preflight_purity", return_value=None), \
+                patch("mcp_server._canonical_repository_slug",
+                      return_value=(None, ["canonical root unresolvable"])), \
+                patch.object(mcp_server, "_preflight_resolved_task",
+                             CANONICAL_TASK), \
+                patch.dict(os.environ, RECONCILER_ENV, clear=True):
+            gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=False, remote="prgs")
+            result = gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=True, remote="prgs")
+        self.assertFalse(result["success"])
+        self.assertEqual(result["blocker_kind"], "repository_binding")
+        self.assertEqual(calls["post"], [])
+
+
+class TestPreflightBinding(_ToolCase):
+    """4. The apply path binds the shared preflight to the exact task."""
+
+    def test_apply_forwards_task_and_target_to_preflight(self):
+        comments = [_lease_comment()]
+        self._run(RECONCILER_ENV, apply=False, comments=comments)
+        side, _calls = _api_side_effect(
+            pr_state="closed", pr_merged=True, comments=comments)
+        seen = {}
+
+        def _purity(remote=None, worktree_path=None, task=None, **kw):
+            seen.update({"remote": remote, "worktree_path": worktree_path,
+                         "task": task, **kw})
+            return None
+
+        with patch("mcp_server.api_request", side_effect=side), \
+                patch("mcp_server.get_auth_header", return_value=FAKE_AUTH), \
+                patch("mcp_server.verify_preflight_purity", _purity), \
+                patch("mcp_server._bound_repository_slug", return_value=SLUG), \
+                patch("mcp_server._repository_binding_block", return_value=None), \
+                patch.object(mcp_server, "_preflight_resolved_task",
+                             CANONICAL_TASK), \
+                patch.dict(os.environ, RECONCILER_ENV, clear=True):
+            result = gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=True, remote="prgs",
+                worktree_path="/tmp/branches/reconciler-745",
+                org="Scaled-Tech-Consulting", repo="Gitea-Tools")
+        self.assertTrue(result["cleanup_performed"])
+        self.assertEqual(seen["task"], CANONICAL_TASK)
+        self.assertEqual(seen["worktree_path"], "/tmp/branches/reconciler-745")
+        self.assertEqual(seen["org"], "Scaled-Tech-Consulting")
+        self.assertEqual(seen["repo"], "Gitea-Tools")
+
+    def test_dry_run_does_not_require_preflight(self):
+        def _boom(*a, **k):
+            raise AssertionError("dry run must not run mutation preflight")
+
+        side, calls = _api_side_effect(
+            pr_state="closed", pr_merged=True, comments=[_lease_comment()])
+        with patch("mcp_server.api_request", side_effect=side), \
+                patch("mcp_server.get_auth_header", return_value=FAKE_AUTH), \
+                patch("mcp_server.verify_preflight_purity", _boom), \
+                patch("mcp_server._bound_repository_slug", return_value=SLUG), \
+                patch.dict(os.environ, AUTHOR_ENV, clear=True):
+            result = gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=False, remote="prgs")
+        self.assertTrue(result["success"])
+        self.assertEqual(calls["post"], [])
+
+
+class TestNoProductionLeaseTouched(unittest.TestCase):
+    """16. No real production lease or PR is referenced by these tests."""
+
+    PRODUCTION_PR = 744
+    PRODUCTION_SESSION = "33673-1d54887a0415"
+    PRODUCTION_MARKER = 12452
+
+    def test_fixtures_are_synthetic(self):
+        self.assertNotEqual(PR, self.PRODUCTION_PR)
+        self.assertNotEqual(SESSION, self.PRODUCTION_SESSION)
+        self.assertNotEqual(LEASE_COMMENT_ID, self.PRODUCTION_MARKER)
+
+    def test_module_source_never_names_the_production_lease(self):
+        source = _Path(__file__).read_text()
+        for token in (self.PRODUCTION_SESSION, str(self.PRODUCTION_MARKER)):
+            self.assertEqual(
+                source.count(token), 1,
+                f"{token!r} must appear only in this guard's own constants",
+            )
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_issue_751_checks_assessor.py

@@ -0,0 +1,602 @@
+"""Regression coverage for the PR checks assessor defect (#751).
+
+Gitea's *combined* commit status reports ``state: pending`` both when a check is
+executing and when the status-context collection is empty. The assessor read
+``state`` alone and defaulted ``checks_required`` to ``True``, so a PR whose head
+had no status contexts — and never would — was routed to ``blocked`` forever.
+
+These tests pin the corrected semantics end to end: live branch protection
+decides whether checks are required, and the actual context collection decides
+what the checks say.
+"""
+
+from __future__ import annotations
+
+import sys
+import unittest
+from unittest.mock import patch
+
+sys.path.insert(0, str(__import__("pathlib").Path(__file__).resolve().parent.parent))
+
+import pr_sync_status  # noqa: E402
+from pr_sync_status import (  # noqa: E402
+    ACTION_BLOCKED,
+    ACTION_MERGE_NOW,
+    ACTION_UPDATE_BRANCH_BY_MERGE,
+    CHECKS_FAILURE,
+    CHECKS_MISSING_REQUIRED,
+    CHECKS_NONE,
+    CHECKS_NOT_REQUIRED,
+    CHECKS_PENDING,
+    CHECKS_SUCCESS,
+    CHECKS_UNKNOWN,
+    assess_pr_sync_status,
+    classify_commit_checks,
+)
+
+
+def _sha(prefix: str) -> str:
+    return (prefix + "0" * 40)[:40]
+
+
+PR_HEAD = _sha("aaaaaaaa")
+BASE_HEAD = _sha("bbbbbbbb")
+
+
+def _base_kwargs(**overrides):
+    data = {
+        "host": "gitea.prgs.cc",
+        "org": "Scaled-Tech-Consulting",
+        "repo": "Gitea-Tools",
+        "pr_number": 751,
+        "pr_state": "open",
+        "source_branch": "fix/issue-751-checks-assessor",
+        "pr_head_sha": PR_HEAD,
+        "base_head_sha": BASE_HEAD,
+        "commits_behind": 0,
+        "mergeable": True,
+        "has_conflicts": False,
+        "branch_protection_requires_current_base": False,
+        "approval_at_current_head": True,
+        "checks_status": "success",
+        "active_author_lock": True,
+        "active_reviewer_lease": False,
+        "active_merger_lease": False,
+    }
+    data.update(overrides)
+    return data
+
+
+def _reasons(result) -> str:
+    return " | ".join(result["reasons"]).lower()
+
+
+class TestClassifyCommitChecks(unittest.TestCase):
+    """Pure classification from live evidence."""
+
+    def test_empty_collection_with_combined_pending_is_not_executing_ci(self):
+        # The exact PR #750 failure mode at the classification layer.
+        result = classify_commit_checks(
+            combined_state="pending",
+            statuses=[],
+            checks_enabled=True,
+            required_contexts=[],
+        )
+        self.assertNotEqual(result["checks_status"], CHECKS_PENDING)
+        self.assertEqual(result["checks_status"], CHECKS_NONE)
+        self.assertEqual(result["context_count"], 0)
+        joined = " ".join(result["reasons"]).lower()
+        self.assertIn("empty", joined)
+        self.assertIn("does not indicate", joined)
+
+    def test_protection_disables_status_checks(self):
+        result = classify_commit_checks(
+            combined_state="pending",
+            statuses=[],
+            checks_enabled=False,
+            required_contexts=[],
+        )
+        self.assertFalse(result["checks_required"])
+        self.assertEqual(result["checks_status"], CHECKS_NOT_REQUIRED)
+
+    def test_no_required_contexts_aggregates_reported_contexts(self):
+        result = classify_commit_checks(
+            combined_state="success",
+            statuses=[{"context": "build", "status": "success"}],
+            checks_enabled=True,
+            required_contexts=[],
+        )
+        self.assertTrue(result["checks_required"])
+        self.assertEqual(result["checks_status"], CHECKS_SUCCESS)
+
+    def test_required_checks_pending(self):
+        result = classify_commit_checks(
+            combined_state="pending",
+            statuses=[{"context": "build", "status": "pending"}],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(result["checks_status"], CHECKS_PENDING)
+
+    def test_required_checks_failed(self):
+        result = classify_commit_checks(
+            combined_state="failure",
+            statuses=[{"context": "build", "status": "failure"}],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(result["checks_status"], CHECKS_FAILURE)
+
+    def test_required_checks_successful(self):
+        result = classify_commit_checks(
+            combined_state="success",
+            statuses=[{"context": "build", "status": "success"}],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(result["checks_status"], CHECKS_SUCCESS)
+        self.assertTrue(result["checks_required"])
+
+    def test_required_context_configured_with_no_matching_result(self):
+        result = classify_commit_checks(
+            combined_state="success",
+            statuses=[{"context": "lint", "status": "success"}],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(result["checks_status"], CHECKS_MISSING_REQUIRED)
+        self.assertEqual(result["missing_required_contexts"], ["build"])
+
+    def test_mixed_required_and_unrelated_contexts_ignores_unrelated(self):
+        # An unrelated failing context must not fail a satisfied required set.
+        result = classify_commit_checks(
+            combined_state="failure",
+            statuses=[
+                {"context": "build", "status": "success"},
+                {"context": "optional-scan", "status": "failure"},
+            ],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(result["checks_status"], CHECKS_SUCCESS)
+        # ...and a failing *required* context still fails despite passing extras.
+        failing = classify_commit_checks(
+            combined_state="success",
+            statuses=[
+                {"context": "build", "status": "failure"},
+                {"context": "optional-scan", "status": "success"},
+            ],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(failing["checks_status"], CHECKS_FAILURE)
+
+    def test_policy_unreadable_fails_closed(self):
+        result = classify_commit_checks(
+            combined_state=None,
+            statuses=[],
+            checks_enabled=None,
+            required_contexts=[],
+            policy_determinable=False,
+        )
+        self.assertTrue(result["checks_required"])
+        self.assertEqual(result["checks_status"], CHECKS_UNKNOWN)
+        self.assertIn("fail closed", " ".join(result["reasons"]).lower())
+
+    def test_malformed_policy_indeterminate_flag_fails_closed(self):
+        result = classify_commit_checks(
+            combined_state="success",
+            statuses=[{"context": "build", "status": "success"}],
+            checks_enabled=None,
+            required_contexts=[],
+            policy_determinable=True,
+        )
+        self.assertTrue(result["checks_required"])
+        self.assertEqual(result["checks_status"], CHECKS_UNKNOWN)
+
+    def test_status_collection_unreadable_fails_closed_when_required(self):
+        result = classify_commit_checks(
+            checks_enabled=True,
+            required_contexts=["build"],
+            status_determinable=False,
+        )
+        self.assertTrue(result["checks_required"])
+        self.assertEqual(result["checks_status"], CHECKS_UNKNOWN)
+
+    def test_unrecognized_context_state_never_reads_as_success(self):
+        result = classify_commit_checks(
+            combined_state="success",
+            statuses=[{"context": "build", "status": "banana"}],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(result["checks_status"], CHECKS_UNKNOWN)
+
+    def test_newest_wins_per_context(self):
+        # Gitea returns newest-first; the stale failure must not win.
+        result = classify_commit_checks(
+            combined_state="success",
+            statuses=[
+                {"context": "build", "status": "success"},
+                {"context": "build", "status": "failure"},
+            ],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(result["checks_status"], CHECKS_SUCCESS)
+
+    def test_malformed_status_rows_are_discarded_not_treated_as_passing(self):
+        result = classify_commit_checks(
+            combined_state="pending",
+            statuses=[{"context": "build"}, "not-a-dict", None],
+            checks_enabled=True,
+            required_contexts=["build"],
+        )
+        self.assertEqual(result["checks_status"], CHECKS_MISSING_REQUIRED)
+
+
+class TestChecksGateSemantics(unittest.TestCase):
+    """``assess_pr_sync_status`` routing and blocker reasons."""
+
+    def test_not_required_allows_merge_now_with_empty_checks(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status=CHECKS_NOT_REQUIRED),
+            checks_required=False,
+        )
+        self.assertEqual(result["recommended_next_action"], ACTION_MERGE_NOW)
+        self.assertTrue(result["approval_valid_for_merge"])
+        self.assertFalse(result["checks_required"])
+        self.assertIn("does not require status checks", _reasons(result))
+
+    def test_none_blocks_when_checks_required(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status=CHECKS_NONE),
+            checks_required=True,
+        )
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+        self.assertIn("no status context", _reasons(result))
+        self.assertIn("not executing ci", _reasons(result))
+
+    def test_missing_required_blocks(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status=CHECKS_MISSING_REQUIRED),
+            checks_required=True,
+        )
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+        self.assertIn("no matching status result", _reasons(result))
+
+    def test_required_pending_blocks(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status=CHECKS_PENDING), checks_required=True
+        )
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+        self.assertIn("not finished", _reasons(result))
+
+    def test_required_failure_blocks(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status=CHECKS_FAILURE), checks_required=True
+        )
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+        self.assertIn("failed", _reasons(result))
+
+    def test_required_success_merges(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status=CHECKS_SUCCESS), checks_required=True
+        )
+        self.assertEqual(result["recommended_next_action"], ACTION_MERGE_NOW)
+
+    def test_unknown_blocks_when_required(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status=CHECKS_UNKNOWN), checks_required=True
+        )
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+        self.assertIn("unknown", _reasons(result))
+
+    def test_unrecognized_status_does_not_fall_through_to_merge(self):
+        # Regression: the previous gate only handled a fixed vocabulary and let
+        # anything else reach merge_now.
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status="totally-unexpected"),
+            checks_required=True,
+        )
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+        self.assertIn("unrecognized checks status", _reasons(result))
+
+    def test_checks_gate_does_not_bypass_approval_requirement(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(
+                checks_status=CHECKS_NOT_REQUIRED, approval_at_current_head=False
+            ),
+            checks_required=False,
+        )
+        self.assertNotEqual(result["recommended_next_action"], ACTION_MERGE_NOW)
+        self.assertFalse(result["approval_valid_for_merge"])
+
+    def test_checks_gate_does_not_bypass_current_base_protection(self):
+        # Existing current-base behavior stays intact (#727).
+        result = assess_pr_sync_status(
+            **_base_kwargs(
+                checks_status=CHECKS_NOT_REQUIRED,
+                commits_behind=3,
+                branch_protection_requires_current_base=True,
+            ),
+            checks_required=False,
+        )
+        self.assertEqual(
+            result["recommended_next_action"], ACTION_UPDATE_BRANCH_BY_MERGE
+        )
+
+    def test_checks_gate_does_not_bypass_conflicts(self):
+        result = assess_pr_sync_status(
+            **_base_kwargs(checks_status=CHECKS_NOT_REQUIRED, mergeable=False),
+            checks_required=False,
+        )
+        self.assertNotEqual(result["recommended_next_action"], ACTION_MERGE_NOW)
+
+
+class TestBranchProtectionPolicy(unittest.TestCase):
+    """Live protection reader derives the status-check requirement."""
+
+    def setUp(self):
+        import gitea_mcp_server as gms
+
+        self.gms = gms
+
+    def _policy(self, responses):
+        def fake_api_request(method, url, auth, *args, **kwargs):
+            for fragment, payload in responses.items():
+                if fragment in url:
+                    if isinstance(payload, Exception):
+                        raise payload
+                    return payload
+            return None
+
+        with patch.object(self.gms, "api_request", side_effect=fake_api_request):
+            return self.gms._branch_protection_policy(
+                "https://example/api/v1/repos/o/r", {"h": "1"}, base_branch="master"
+            )
+
+    def test_status_checks_enabled_with_contexts(self):
+        policy = self._policy({
+            "branch_protections": [{
+                "branch_name": "master",
+                "block_on_outdated_branch": True,
+                "enable_status_check": True,
+                "status_check_contexts": ["ci/build", " "],
+            }],
+        })
+        self.assertTrue(policy["determinable"])
+        self.assertTrue(policy["checks_enabled"])
+        self.assertTrue(policy["requires_current_base"])
+        self.assertEqual(policy["required_contexts"], ["ci/build"])
+
+    def test_status_checks_disabled(self):
+        policy = self._policy({
+            "branch_protections": [{
+                "branch_name": "master",
+                "enable_status_check": False,
+            }],
+        })
+        self.assertTrue(policy["determinable"])
+        self.assertFalse(policy["checks_enabled"])
+
+    def test_no_protection_rule_means_checks_not_required(self):
+        # PR #750's repository shape: protection list readable but empty.
+        policy = self._policy({"branch_protections": [], "branches/master": {}})
+        self.assertTrue(policy["determinable"])
+        self.assertFalse(policy["protection_found"])
+        self.assertFalse(policy["checks_enabled"])
+        self.assertIsNone(policy["requires_current_base"])
+
+    def test_protection_without_status_field_means_not_enabled(self):
+        policy = self._policy({
+            "branch_protections": [{
+                "branch_name": "master",
+                "block_on_outdated_branch": True,
+            }],
+        })
+        self.assertTrue(policy["determinable"])
+        self.assertFalse(policy["checks_enabled"])
+        self.assertTrue(policy["requires_current_base"])
+
+    def test_api_failure_is_not_determinable(self):
+        policy = self._policy({
+            "branch_protections": RuntimeError("boom"),
+        })
+        self.assertFalse(policy["determinable"])
+        self.assertIsNone(policy["checks_enabled"])
+
+    def test_missing_branch_is_not_determinable(self):
+        with patch.object(self.gms, "api_request", return_value=[]):
+            policy = self.gms._branch_protection_policy(
+                "https://example/api/v1/repos/o/r", {}, base_branch="  "
+            )
+        self.assertFalse(policy["determinable"])
+
+    def test_requires_current_base_accessor_preserved(self):
+        def fake(method, url, auth, *a, **k):
+            if "branch_protections" in url:
+                return [{"branch_name": "master",
+                         "block_on_outdated_branch": True}]
+            return None
+
+        with patch.object(self.gms, "api_request", side_effect=fake):
+            value = self.gms._branch_protection_requires_current_base(
+                "https://example/api/v1/repos/o/r", {}, base_branch="master"
+            )
+        self.assertTrue(value)
+
+
+class TestCommitChecksSnapshot(unittest.TestCase):
+    def setUp(self):
+        import gitea_mcp_server as gms
+
+        self.gms = gms
+
+    def test_reads_state_and_context_collection(self):
+        payload = {"state": "pending", "statuses": [{"context": "b",
+                                                     "status": "pending"}]}
+        with patch.object(self.gms, "api_request", return_value=payload):
+            snap = self.gms._commit_checks_snapshot(
+                "https://example/api/v1/repos/o/r", {}, sha=PR_HEAD
+            )
+        self.assertTrue(snap["determinable"])
+        self.assertEqual(snap["combined_state"], "pending")
+        self.assertEqual(len(snap["statuses"]), 1)
+
+    def test_empty_collection_recorded_as_determinable(self):
+        with patch.object(
+            self.gms, "api_request", return_value={"state": "pending", "statuses": []}
+        ):
+            snap = self.gms._commit_checks_snapshot(
+                "https://example/api/v1/repos/o/r", {}, sha=PR_HEAD
+            )
+        self.assertTrue(snap["determinable"])
+        self.assertEqual(snap["statuses"], [])
+
+    def test_api_failure_is_not_determinable(self):
+        with patch.object(self.gms, "api_request", side_effect=RuntimeError("x")):
+            snap = self.gms._commit_checks_snapshot(
+                "https://example/api/v1/repos/o/r", {}, sha=PR_HEAD
+            )
+        self.assertFalse(snap["determinable"])
+
+
+class TestMcpWrapperForwarding(unittest.TestCase):
+    """The production MCP path must reach the derived ``checks_required``."""
+
+    def setUp(self):
+        import gitea_mcp_server as gms
+
+        self.gms = gms
+        self.base = (
+            "https://gitea.prgs.cc/api/v1/repos/Scaled-Tech-Consulting/Gitea-Tools"
+        )
+
+    def _patches(self, fake_api_request, spy):
+        return [
+            patch.object(self.gms, "_profile_operation_gate", return_value=None),
+            patch.object(self.gms, "_resolve", return_value=(
+                "gitea.prgs.cc", "Scaled-Tech-Consulting", "Gitea-Tools")),
+            patch.object(self.gms, "_auth", return_value={"Authorization": "x"}),
+            patch.object(self.gms, "repo_api_url", return_value=self.base),
+            patch.object(self.gms, "api_request", side_effect=fake_api_request),
+            patch.object(self.gms, "gitea_get_pr_review_feedback", return_value={
+                "success": True, "approval_at_current_head": True}),
+            patch.object(self.gms, "_prove_author_ownership_for_pr", return_value={
+                "has_author_lock": True}),
+            patch.object(pr_sync_status, "assess_pr_sync_status", side_effect=spy),
+        ]
+
+    def _run(self, *, protections, status_payload, pr_number=750,
+             caller_checks_status=None):
+        captured = {}
+        real_assess = pr_sync_status.assess_pr_sync_status
+
+        def spy(**kwargs):
+            captured.update(kwargs)
+            return real_assess(**kwargs)
+
+        def fake_api_request(method, url, auth, *args, **kwargs):
+            if f"/pulls/{pr_number}" in url:
+                return {
+                    "number": pr_number,
+                    "state": "open",
+                    "title": "t",
+                    "body": "b",
+                    "mergeable": True,
+                    "head": {"sha": PR_HEAD, "ref": "fix/x"},
+                    "base": {"sha": BASE_HEAD, "ref": "master"},
+                }
+            if "/branch_protections" in url:
+                if isinstance(protections, Exception):
+                    raise protections
+                return protections
+            if "/branches/master" in url:
+                return {"commit": {"id": BASE_HEAD}}
+            if "/status" in url:
+                if isinstance(status_payload, Exception):
+                    raise status_payload
+                return status_payload
+            if "/compare/" in url:
+                return {"total_commits": 0}
+            if "/comments" in url:
+                return []
+            return None
+
+        stack = self._patches(fake_api_request, spy)
+        for p in stack:
+            p.start()
+        try:
+            result = self.gms.gitea_assess_pr_sync_status(
+                pr_number=pr_number, remote="prgs",
+                checks_status=caller_checks_status,
+            )
+        finally:
+            for p in reversed(stack):
+                p.stop()
+        return result, captured
+
+    def test_derived_checks_required_is_forwarded(self):
+        result, captured = self._run(
+            protections=[{"branch_name": "master", "enable_status_check": True,
+                          "status_check_contexts": ["ci/build"]}],
+            status_payload={"state": "pending", "statuses": [
+                {"context": "ci/build", "status": "pending"}]},
+        )
+        self.assertIn("checks_required", captured)
+        self.assertTrue(captured["checks_required"])
+        self.assertEqual(captured["checks_status"], CHECKS_PENDING)
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+
+    def test_pr750_empty_status_with_no_protection_is_merge_ready(self):
+        # The exact reported failure: combined pending, zero contexts, and no
+        # branch protection requiring checks.
+        result, captured = self._run(
+            protections=[],
+            status_payload={"state": "pending", "statuses": []},
+        )
+        self.assertFalse(captured["checks_required"])
+        self.assertEqual(captured["checks_status"], CHECKS_NOT_REQUIRED)
+        self.assertNotEqual(result["checks_status"], CHECKS_PENDING)
+        self.assertEqual(result["recommended_next_action"], ACTION_MERGE_NOW)
+
+    def test_protection_read_failure_blocks(self):
+        result, captured = self._run(
+            protections=RuntimeError("protection unavailable"),
+            status_payload={"state": "success", "statuses": []},
+        )
+        self.assertTrue(captured["checks_required"])
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+
+    def test_caller_supplied_status_cannot_mask_live_required_failure(self):
+        # A caller-declared "success" must not override live evidence.
+        result, captured = self._run(
+            protections=[{"branch_name": "master", "enable_status_check": True,
+                          "status_check_contexts": ["ci/build"]}],
+            status_payload={"state": "failure", "statuses": [
+                {"context": "ci/build", "status": "failure"}]},
+            caller_checks_status="success",
+        )
+        self.assertEqual(captured["checks_status"], CHECKS_FAILURE)
+        self.assertEqual(result["recommended_next_action"], ACTION_BLOCKED)
+        self.assertEqual(
+            result["checks_evidence"]["caller_supplied_checks_status"], "success"
+        )
+
+    def test_checks_evidence_is_reported_without_secrets(self):
+        result, _ = self._run(
+            protections=[],
+            status_payload={"state": "pending", "statuses": []},
+        )
+        evidence = result["checks_evidence"]
+        self.assertEqual(evidence["context_count"], 0)
+        self.assertEqual(evidence["combined_state"], "pending")
+        self.assertFalse(evidence["protection_found"])
+        blob = repr(result).lower()
+        self.assertNotIn("authorization", blob)
+        self.assertNotIn("token", blob)
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_issue_753_dead_pid_lock_recovery.py

@@ -0,0 +1,364 @@
+"""Dead-session author issue-lock recovery (#753).
+
+Covers the narrow recovery path that lets an author re-acquire a durable lock
+after the MCP session that recorded it exits, plus every rejection condition
+that must keep failing closed.
+"""
+
+import os
+import subprocess
+import sys
+import unittest
+from datetime import datetime, timedelta, timezone
+
+sys.path.insert(0, str(__import__("pathlib").Path(__file__).resolve().parent.parent))
+
+import issue_lock_recovery  # noqa: E402
+import issue_lock_store  # noqa: E402
+import issue_lock_worktree  # noqa: E402
+
+ISSUE = 4242
+BRANCH = f"fix/issue-{ISSUE}-demo"
+WORKTREE = "/scratch/wt"
+HEAD = "a" * 40
+OTHER_SHA = "b" * 40
+IDENTITY = "example-user"
+PROFILE = "example-author"
+
+
+def dead_pid() -> int:
+    """A PID that has certainly exited (spawned, then reaped)."""
+    proc = subprocess.Popen([sys.executable, "-c", "pass"])
+    proc.wait()
+    return proc.pid
+
+
+def future_ts(hours: int = 4) -> str:
+    return (
+        (datetime.now(timezone.utc) + timedelta(hours=hours))
+        .isoformat()
+        .replace("+00:00", "Z")
+    )
+
+
+def make_lock(**overrides):
+    lock = {
+        "issue_number": ISSUE,
+        "branch_name": BRANCH,
+        "worktree_path": WORKTREE,
+        "remote": "prgs",
+        "org": "ExampleOrg",
+        "repo": "ExampleRepo",
+        "session_pid": dead_pid(),
+        "work_lease": {
+            "operation_type": issue_lock_store.AUTHOR_ISSUE_WORK_LEASE,
+            "issue_number": ISSUE,
+            "branch": BRANCH,
+            "worktree_path": WORKTREE,
+            "claimant": {"username": IDENTITY, "profile": PROFILE},
+            "expires_at": future_ts(),
+        },
+    }
+    lock.update(overrides)
+    return lock
+
+
+def assess(lock=None, **overrides):
+    kwargs = {
+        "issue_number": ISSUE,
+        "branch_name": BRANCH,
+        "worktree_path": WORKTREE,
+        "remote": "prgs",
+        "org": "ExampleOrg",
+        "repo": "ExampleRepo",
+        "identity": IDENTITY,
+        "profile": PROFILE,
+        "current_branch": BRANCH,
+        "porcelain_status": "",
+        "head_sha": HEAD,
+        "remote_head_sha": HEAD,
+        "pr_head_sha": HEAD,
+        "pr_number": 99,
+        "competing_live_locks": [],
+        "candidate_branches": [BRANCH],
+        "current_pid": os.getpid(),
+    }
+    kwargs.update(overrides)
+    return issue_lock_recovery.assess_dead_session_lock_recovery(
+        make_lock() if lock is None else lock, **kwargs
+    )
+
+
+class TestDeadSessionRecoveryGranted(unittest.TestCase):
+    def test_dead_pid_with_exact_evidence_recovers(self):
+        result = assess()
+        self.assertTrue(result["recovery_sanctioned"], result["reasons"])
+        self.assertEqual(result["outcome"], issue_lock_recovery.RECOVERY_SANCTIONED)
+
+    def test_recovery_still_granted_when_no_open_pr_exists(self):
+        # A locked branch need not have a PR yet; absence must not block.
+        result = assess(pr_head_sha=None, pr_number=None)
+        self.assertTrue(result["recovery_sanctioned"], result["reasons"])
+
+    def test_lease_expiry_is_not_required_for_recovery(self):
+        # The defining condition is PID death, not TTL expiry (the #601 gap).
+        lock = make_lock()
+        self.assertFalse(issue_lock_store.is_lease_expired(lock))
+        self.assertFalse(issue_lock_store.assess_lock_freshness(lock)["live"])
+        self.assertTrue(assess(lock)["recovery_sanctioned"])
+
+
+class TestDeadSessionRecoveryRefused(unittest.TestCase):
+    def assert_refused(self, result, needle):
+        self.assertFalse(result["recovery_sanctioned"])
+        self.assertEqual(result["outcome"], issue_lock_recovery.REFUSED)
+        self.assertTrue(
+            any(needle in reason for reason in result["reasons"]),
+            f"expected {needle!r} in {result['reasons']}",
+        )
+
+    def test_live_prior_pid_refused(self):
+        lock = make_lock(session_pid=os.getpid(), pid=os.getpid())
+        # Distinct current pid so the refusal is attributable to liveness.
+        self.assert_refused(assess(lock, current_pid=os.getpid() + 1), "still alive")
+
+    def test_different_author_identity_refused(self):
+        self.assert_refused(
+            assess(identity="someone-else"), "does not match active identity"
+        )
+
+    def test_different_profile_refused(self):
+        self.assert_refused(
+            assess(profile="other-profile"), "does not match active profile"
+        )
+
+    def test_different_branch_refused(self):
+        lock = make_lock(branch_name=f"fix/issue-{ISSUE}-other")
+        self.assert_refused(assess(lock), "does not match requested")
+
+    def test_worktree_parked_on_another_branch_refused(self):
+        self.assert_refused(assess(current_branch="master"), "not the locked branch")
+
+    def test_detached_head_worktree_refused(self):
+        self.assert_refused(assess(current_branch=None), "detached HEAD")
+
+    def test_different_worktree_refused(self):
+        self.assert_refused(
+            assess(worktree_path="/scratch/elsewhere"), "does not match declared"
+        )
+
+    def test_dirty_worktree_refused(self):
+        self.assert_refused(
+            assess(porcelain_status=" M gitea_mcp_server.py\n"), "requires a clean"
+        )
+
+    def test_local_head_differing_from_remote_refused(self):
+        self.assert_refused(
+            assess(remote_head_sha=OTHER_SHA), "does not match remote branch head"
+        )
+
+    def test_pr_head_differing_refused(self):
+        self.assert_refused(assess(pr_head_sha=OTHER_SHA), "does not match local head")
+
+    def test_missing_remote_head_refused(self):
+        self.assert_refused(assess(remote_head_sha=None), "remote head")
+
+    def test_competing_live_lock_refused(self):
+        competing = [
+            {
+                "issue_number": ISSUE,
+                "branch_name": BRANCH,
+                "worktree_path": "/scratch/other-wt",
+                "pid": os.getpid(),
+            }
+        ]
+        self.assert_refused(
+            assess(competing_live_locks=competing), "competing live lock"
+        )
+
+    def test_unrelated_live_lock_does_not_block(self):
+        unrelated = [
+            {
+                "issue_number": 999,
+                "branch_name": "fix/issue-999-unrelated",
+                "worktree_path": "/scratch/unrelated",
+                "pid": os.getpid(),
+            }
+        ]
+        self.assertTrue(assess(competing_live_locks=unrelated)["recovery_sanctioned"])
+
+    def test_multiple_candidate_branches_refused(self):
+        self.assert_refused(
+            assess(candidate_branches=[BRANCH, f"feat/issue-{ISSUE}-rival"]),
+            "multiple branches claim this issue",
+        )
+
+    def test_repository_scope_mismatch_refused(self):
+        self.assert_refused(assess(repo="OtherRepo"), "does not match requested")
+
+    def test_malformed_lock_missing_worktree_refused(self):
+        lock = make_lock()
+        lock.pop("worktree_path")
+        self.assert_refused(assess(lock), "incomplete")
+
+    def test_malformed_lock_missing_pid_refused(self):
+        lock = make_lock()
+        lock.pop("session_pid", None)
+        lock.pop("pid", None)
+        self.assert_refused(assess(lock), "incomplete")
+
+    def test_lock_without_claimant_refused(self):
+        lock = make_lock()
+        lock["work_lease"] = dict(lock["work_lease"])
+        lock["work_lease"].pop("claimant")
+        self.assert_refused(assess(lock), "claimant identity/profile")
+
+
+class TestNotACandidate(unittest.TestCase):
+    def test_absent_lock_is_not_a_candidate(self):
+        result = issue_lock_recovery.assess_dead_session_lock_recovery(
+            None,
+            issue_number=ISSUE,
+            branch_name=BRANCH,
+            worktree_path=WORKTREE,
+            remote="prgs",
+            org="ExampleOrg",
+            repo="ExampleRepo",
+            identity=IDENTITY,
+            profile=PROFILE,
+            current_branch=BRANCH,
+            porcelain_status="",
+            head_sha=HEAD,
+            remote_head_sha=HEAD,
+        )
+        self.assertEqual(result["outcome"], issue_lock_recovery.NO_CANDIDATE)
+        self.assertFalse(result["recovery_sanctioned"])
+        self.assertFalse(result["is_candidate"])
+
+    def test_lock_for_a_different_issue_is_not_a_candidate(self):
+        result = assess(make_lock(issue_number=7777))
+        self.assertEqual(result["outcome"], issue_lock_recovery.NO_CANDIDATE)
+        self.assertFalse(result["recovery_sanctioned"])
+
+
+class TestWorktreeGateWaiver(unittest.TestCase):
+    def test_new_issue_claim_still_requires_base_equivalence(self):
+        result = issue_lock_worktree.assess_issue_lock_worktree(
+            worktree_path=WORKTREE,
+            current_branch=BRANCH,
+            porcelain_status="",
+            base_equivalent=False,
+        )
+        self.assertTrue(result["block"])
+        self.assertFalse(result["base_equivalence_waived"])
+
+    def test_sanctioned_recovery_waives_base_equivalence(self):
+        result = issue_lock_worktree.assess_issue_lock_worktree(
+            worktree_path=WORKTREE,
+            current_branch=BRANCH,
+            porcelain_status="",
+            base_equivalent=False,
+            recovery_sanctioned=True,
+        )
+        self.assertTrue(result["proven"], result["reasons"])
+        self.assertTrue(result["base_equivalence_waived"])
+
+    def test_recovery_never_waives_cleanliness(self):
+        result = issue_lock_worktree.assess_issue_lock_worktree(
+            worktree_path=WORKTREE,
+            current_branch=BRANCH,
+            porcelain_status=" M gitea_mcp_server.py\n",
+            base_equivalent=False,
+            recovery_sanctioned=True,
+        )
+        self.assertTrue(result["block"])
+        self.assertTrue(
+            any("tracked file edits" in reason for reason in result["reasons"])
+        )
+
+    def test_unproven_base_equivalence_still_blocks_without_recovery(self):
+        result = issue_lock_worktree.assess_issue_lock_worktree(
+            worktree_path=WORKTREE,
+            current_branch=BRANCH,
+            porcelain_status="",
+            base_equivalent=None,
+        )
+        self.assertTrue(result["block"])
+
+
+class TestRecoveryRecordAndDownstream(unittest.TestCase):
+    def test_recovery_record_preserves_truthful_provenance(self):
+        assessment = assess()
+        prior = assessment["evidence"]["prior_session_pid"]
+        record = issue_lock_recovery.build_recovery_record(
+            assessment, recovered_at="2026-07-18T23:21:40Z"
+        )
+        self.assertTrue(record["recovered"])
+        self.assertEqual(record["prior_session_pid"], prior)
+        self.assertEqual(record["replacement_session_pid"], os.getpid())
+        self.assertNotEqual(
+            record["prior_session_pid"], record["replacement_session_pid"]
+        )
+        self.assertFalse(record["prior_pid_alive"])
+        self.assertEqual(record["recovered_at"], "2026-07-18T23:21:40Z")
+        self.assertEqual(record["branch_name"], BRANCH)
+        self.assertEqual(record["local_head"], HEAD)
+        self.assertEqual(record["identity"], IDENTITY)
+        self.assertTrue(record["proof"])
+
+    def test_recovery_record_carries_no_secret_material(self):
+        record = issue_lock_recovery.build_recovery_record(
+            assess(), recovered_at="2026-07-18T23:21:40Z"
+        )
+        blob = repr(record).lower()
+        for banned in ("token", "password", "authorization", "secret", "api_key"):
+            self.assertNotIn(banned, blob)
+
+    def test_recovered_lock_satisfies_update_by_merge_ownership(self):
+        # After recovery the lock is rebound to the live session, so the
+        # ownership re-check used by gitea_update_pr_branch_by_merge passes.
+        assessment = assess()
+        recovered_lock = make_lock(session_pid=os.getpid(), pid=os.getpid())
+        recovered_lock["dead_session_recovery"] = (
+            issue_lock_recovery.build_recovery_record(
+                assessment, recovered_at="2026-07-18T23:21:40Z"
+            )
+        )
+        freshness = issue_lock_store.assess_lock_freshness(recovered_lock)
+        self.assertTrue(freshness["live"], freshness)
+
+        verdict = issue_lock_store.verify_lock_for_mutation(
+            recovered_lock,
+            issue_number=ISSUE,
+            branch_name=BRANCH,
+            worktree_path=WORKTREE,
+        )
+        self.assertTrue(verdict["proven"], verdict["reasons"])
+        self.assertFalse(verdict["block"])
+
+    def test_pre_recovery_lock_fails_ownership_check(self):
+        # Guards against a false positive above: the dead-PID lock must fail.
+        verdict = issue_lock_store.verify_lock_for_mutation(
+            make_lock(),
+            issue_number=ISSUE,
+            branch_name=BRANCH,
+            worktree_path=WORKTREE,
+        )
+        self.assertTrue(verdict["block"])
+        self.assertTrue(any("not live" in reason for reason in verdict["reasons"]))
+
+    def test_no_manual_file_seeding_required(self):
+        # The whole decision is reachable from the durable record plus live
+        # observation; nothing is written to disk to reach a verdict.
+        self.assertTrue(assess()["recovery_sanctioned"])
+
+    def test_refusal_message_is_fail_closed(self):
+        message = issue_lock_recovery.format_recovery_refusal(
+            assess(porcelain_status=" M x.py\n")
+        )
+        self.assertIn("fail closed", message)
+        self.assertIn("recovery refused", message.lower())
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_issue_755_owning_pr_recovery.py

@@ -0,0 +1,609 @@
+import sys as _sys
+from pathlib import Path as _Path
+_sys.path.insert(0, str(_Path(__file__).resolve().parent))
+from mutation_profile_fixture import shared_mutation_env  # noqa: E402
+"""Dead-session lock recovery when the issue already owns an open PR (#755).
+
+#753 added the recovery *assessor*, but the production ``gitea_lock_issue``
+path still rejected every sanctioned recovery: a dead-session lock is by
+construction a lock for work that already has an open PR, and the #400
+duplicate-work gate blocked unconditionally on any linked open PR. These tests
+drive the real MCP handler, not just the pure assessor, so that gap cannot
+reopen.
+"""
+import os
+import subprocess
+import sys
+import tempfile
+import unittest
+from datetime import datetime, timedelta, timezone
+from pathlib import Path
+from unittest.mock import patch
+
+sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
+
+import issue_lock_provenance  # noqa: E402
+import issue_lock_recovery  # noqa: E402
+import issue_lock_store  # noqa: E402
+import mcp_server  # noqa: E402
+from issue_work_duplicate_gate import (  # noqa: E402
+    OUTCOME_DUPLICATE_PR_PREVENTED,
+    OUTCOME_DUPLICATE_WORK_NOT_PREVENTED,
+    PHASE_LOCK,
+    assess_work_issue_duplicate_gate,
+)
+
+ISSUE = 4755
+BRANCH = f"fix/issue-{ISSUE}-owning-pr"
+OTHER_BRANCH = f"fix/issue-{ISSUE}-competing"
+HEAD = "c" * 40
+OTHER_HEAD = "d" * 40
+OWNING_PR = 4756
+OTHER_PR = 4757
+IDENTITY = "example-user"
+PROFILE = "test-author-prgs"
+ORG = "Scaled-Tech-Consulting"
+REPO = "Gitea-Tools"
+
+
+def dead_pid() -> int:
+    """A PID that has certainly exited (spawned, then reaped)."""
+    proc = subprocess.Popen([sys.executable, "-c", "pass"])
+    proc.wait()
+    return proc.pid
+
+
+def shifted_ts(hours: int = 4) -> str:
+    return (
+        (datetime.now(timezone.utc) + timedelta(hours=hours))
+        .isoformat()
+        .replace("+00:00", "Z")
+    )
+
+
+def owning_pr(number=OWNING_PR, ref=BRANCH, sha=HEAD, issue=ISSUE):
+    return {
+        "number": number,
+        "title": f"fix: something (Closes #{issue})",
+        "body": f"Closes #{issue}.",
+        "head": {"ref": ref, "sha": sha},
+    }
+
+
+def sanctioned_token(
+    issue_number=ISSUE, pr_number=OWNING_PR, branch=BRANCH, head=HEAD
+):
+    """The evidence shape the server derives from a granted recovery."""
+    return {
+        "issue_number": issue_number,
+        "pr_number": pr_number,
+        "branch_name": branch,
+        "head_sha": head,
+    }
+
+
+# ───────────────────────── duplicate gate: exemption ─────────────────────────
+
+
+class TestOwningPrExemptionGranted(unittest.TestCase):
+    def test_exact_owning_pr_is_not_duplicate_work(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[owning_pr()],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        self.assertFalse(result["block"])
+        self.assertTrue(result["owning_pr_recovery_exempted"])
+        self.assertEqual(result["outcome"], OUTCOME_DUPLICATE_WORK_NOT_PREVENTED)
+        self.assertEqual(result["linked_open_pr"], OWNING_PR)
+        self.assertEqual(result["linked_open_pr_count"], 1)
+
+    def test_unrelated_open_pr_alongside_owning_pr_is_ignored(self):
+        unrelated = {
+            "number": 999,
+            "title": "chore: unrelated",
+            "body": "no linkage",
+            "head": {"ref": "chore/unrelated", "sha": OTHER_HEAD},
+        }
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[unrelated, owning_pr()],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        self.assertFalse(result["block"])
+        self.assertTrue(result["owning_pr_recovery_exempted"])
+        self.assertEqual(result["linked_open_pr_count"], 1)
+
+
+class TestOwningPrExemptionRefused(unittest.TestCase):
+    def assert_blocked(self, result):
+        self.assertTrue(result["block"])
+        self.assertFalse(result["owning_pr_recovery_exempted"])
+        self.assertEqual(result["outcome"], OUTCOME_DUPLICATE_PR_PREVENTED)
+
+    def test_no_recovery_evidence_keeps_ordinary_blocker(self):
+        self.assert_blocked(
+            assess_work_issue_duplicate_gate(
+                ISSUE,
+                open_prs=[owning_pr()],
+                branch_names=[BRANCH],
+                claim_entry={"status": "not_claimed"},
+                locked_branch=BRANCH,
+                phase=PHASE_LOCK,
+            )
+        )
+
+    def test_competing_pr_number_refused(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[owning_pr(number=OTHER_PR)],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        self.assert_blocked(result)
+
+    def test_multiple_linked_open_prs_refused(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[owning_pr(), owning_pr(number=OTHER_PR, ref=OTHER_BRANCH)],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        self.assert_blocked(result)
+        self.assertEqual(result["linked_open_pr_count"], 2)
+
+    def test_different_branch_refused(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[owning_pr(ref=OTHER_BRANCH)],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        self.assert_blocked(result)
+
+    def test_locked_branch_differing_from_evidence_refused(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[owning_pr()],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=OTHER_BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        self.assert_blocked(result)
+
+    def test_different_head_refused(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[owning_pr(sha=OTHER_HEAD)],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        self.assert_blocked(result)
+
+    def test_evidence_for_another_issue_refused(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[owning_pr()],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(issue_number=ISSUE + 1),
+        )
+        self.assert_blocked(result)
+
+    def test_missing_head_in_live_pr_refused(self):
+        pr = owning_pr()
+        pr["head"] = {"ref": BRANCH}
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[pr],
+            branch_names=[BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        self.assert_blocked(result)
+
+
+class TestOrdinaryDuplicateBehaviorUnchanged(unittest.TestCase):
+    def test_clean_issue_still_passes(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[],
+            branch_names=["feat/other-issue-99"],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+        )
+        self.assertFalse(result["block"])
+        self.assertFalse(result["owning_pr_recovery_exempted"])
+
+    def test_competing_branch_still_blocks_even_with_owning_pr_evidence(self):
+        result = assess_work_issue_duplicate_gate(
+            ISSUE,
+            open_prs=[owning_pr()],
+            branch_names=[BRANCH, OTHER_BRANCH],
+            claim_entry={"status": "not_claimed"},
+            locked_branch=BRANCH,
+            phase=PHASE_LOCK,
+            recovered_owning_pr=sanctioned_token(),
+        )
+        # The owning PR is exempt, but the competing branch is not.
+        self.assertTrue(result["block"])
+        self.assertTrue(result["owning_pr_recovery_exempted"])
+        self.assertIn(OTHER_BRANCH, result["conflicting_branches"])
+
+
+# ─────────────────── server-derived evidence cannot be forged ───────────────────
+
+
+class TestOwningPrEvidenceDerivation(unittest.TestCase):
+    def granted(self, **evidence_overrides):
+        evidence = {
+            "issue_number": ISSUE,
+            "locked_branch": BRANCH,
+            "local_head": HEAD,
+            "remote_head": HEAD,
+            "pr_head": HEAD,
+            "pr_number": OWNING_PR,
+        }
+        evidence.update(evidence_overrides)
+        return {
+            "outcome": issue_lock_recovery.RECOVERY_SANCTIONED,
+            "recovery_sanctioned": True,
+            "is_candidate": True,
+            "reasons": [],
+            "evidence": evidence,
+        }
+
+    def test_granted_recovery_yields_evidence(self):
+        token = issue_lock_recovery.owning_pr_recovery_evidence(self.granted())
+        self.assertEqual(token, sanctioned_token())
+
+    def test_none_assessment_yields_nothing(self):
+        self.assertIsNone(issue_lock_recovery.owning_pr_recovery_evidence(None))
+
+    def test_refused_assessment_yields_nothing(self):
+        refused = self.granted()
+        refused["outcome"] = issue_lock_recovery.REFUSED
+        refused["recovery_sanctioned"] = False
+        self.assertIsNone(issue_lock_recovery.owning_pr_recovery_evidence(refused))
+
+    def test_sanctioned_flag_without_outcome_yields_nothing(self):
+        forged = self.granted()
+        forged["outcome"] = "SOMETHING_ELSE"
+        self.assertIsNone(issue_lock_recovery.owning_pr_recovery_evidence(forged))
+
+    def test_missing_pr_number_yields_nothing(self):
+        self.assertIsNone(
+            issue_lock_recovery.owning_pr_recovery_evidence(
+                self.granted(pr_number=None)
+            )
+        )
+
+    def test_head_disagreement_in_evidence_yields_nothing(self):
+        self.assertIsNone(
+            issue_lock_recovery.owning_pr_recovery_evidence(
+                self.granted(remote_head=OTHER_HEAD)
+            )
+        )
+        self.assertIsNone(
+            issue_lock_recovery.owning_pr_recovery_evidence(
+                self.granted(local_head=OTHER_HEAD)
+            )
+        )
+
+    def test_real_refused_assessment_yields_nothing(self):
+        """End-to-end against the real assessor, not a hand-built dict."""
+        lock = {
+            "issue_number": ISSUE,
+            "branch_name": BRANCH,
+            "worktree_path": "/scratch/wt",
+            "remote": "prgs",
+            "org": "Example-Org",
+            "repo": "Example-Repo",
+            "session_pid": os.getpid(),  # alive → must refuse
+            "work_lease": {
+                "operation_type": issue_lock_store.AUTHOR_ISSUE_WORK_LEASE,
+                "issue_number": ISSUE,
+                "branch": BRANCH,
+                "worktree_path": "/scratch/wt",
+                "claimant": {"username": IDENTITY, "profile": PROFILE},
+                "expires_at": shifted_ts(),
+            },
+        }
+        assessment = issue_lock_recovery.assess_dead_session_lock_recovery(
+            lock,
+            issue_number=ISSUE,
+            branch_name=BRANCH,
+            worktree_path="/scratch/wt",
+            remote="prgs",
+            org="Example-Org",
+            repo="Example-Repo",
+            identity=IDENTITY,
+            profile=PROFILE,
+            current_branch=BRANCH,
+            porcelain_status="",
+            head_sha=HEAD,
+            remote_head_sha=HEAD,
+            pr_head_sha=HEAD,
+            pr_number=OWNING_PR,
+            competing_live_locks=[],
+            candidate_branches=[BRANCH],
+            current_pid=os.getpid(),
+        )
+        self.assertFalse(assessment["recovery_sanctioned"])
+        self.assertIsNone(
+            issue_lock_recovery.owning_pr_recovery_evidence(assessment)
+        )
+
+
+# ──────────────────── end-to-end: the real gitea_lock_issue ────────────────────
+
+
+class LockIssueEndToEndBase(unittest.TestCase):
+    """Drives ``mcp_server.gitea_lock_issue`` with live git/Gitea observation
+    stubbed at the module boundary — the production gate chain itself runs."""
+
+    def setUp(self):
+        self.lock_dir = tempfile.TemporaryDirectory()
+        self.addCleanup(self.lock_dir.cleanup)
+        self.worktree = os.path.realpath(os.getcwd())
+        # Bind host/org/repo to what the ``test-author-prgs`` fixture profile is
+        # pinned to, so the session-context gate under test is the real one and
+        # not a cross-host denial. The issue number and lock dir stay synthetic.
+        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 write_durable_lock(self, *, pid, branch=BRANCH, worktree=None):
+        path = issue_lock_store.lock_file_path(
+            remote="prgs",
+            org=ORG,
+            repo=REPO,
+            issue_number=ISSUE,
+            lock_dir=self.lock_dir.name,
+        )
+        claimant = {"username": IDENTITY, "profile": PROFILE}
+        data = {
+            "issue_number": ISSUE,
+            "branch_name": branch,
+            "remote": "prgs",
+            "org": ORG,
+            "repo": REPO,
+            "worktree_path": worktree or self.worktree,
+            "session_pid": pid,
+            "pid": pid,
+            "work_lease": {
+                "operation_type": issue_lock_store.AUTHOR_ISSUE_WORK_LEASE,
+                "issue_number": ISSUE,
+                "branch": branch,
+                "worktree_path": worktree or self.worktree,
+                "claimant": claimant,
+                "created_at": shifted_ts(-1),
+                "last_heartbeat_at": shifted_ts(-1),
+                "expires_at": shifted_ts(),
+            },
+            "lock_provenance": issue_lock_provenance.build_sanctioned_lock_provenance(
+                tool="gitea_lock_issue",
+                claimant=claimant,
+            ),
+        }
+        issue_lock_store.save_lock_file(path, data)
+        return path
+
+    def run_lock(
+        self,
+        *,
+        open_prs,
+        porcelain="",
+        current_branch=BRANCH,
+        base_equivalent=False,
+        branch_names=None,
+        head_sha=HEAD,
+        remote_head=HEAD,
+    ):
+        branch_names = branch_names if branch_names is not None else [BRANCH]
+        branch_entries = [
+            {"name": name, "commit": {"id": remote_head}} for name in branch_names
+        ]
+        env = shared_mutation_env(
+            "test-author-prgs",
+            include_example_repo=True,
+            GITEA_ISSUE_LOCK_DIR=self.lock_dir.name,
+        )
+        with patch(
+            "mcp_server.api_get_all", return_value=branch_entries
+        ), patch(
+            "mcp_server._list_open_pulls", return_value=list(open_prs)
+        ), patch(
+            "mcp_server.get_auth_header", return_value="token x"
+        ), patch(
+            "mcp_server._work_lease_claimant",
+            return_value={"username": IDENTITY, "profile": PROFILE},
+        ), patch(
+            "mcp_server.issue_lock_worktree.read_worktree_git_state",
+            return_value={
+                "current_branch": current_branch,
+                "porcelain_status": porcelain,
+                "base_equivalent": base_equivalent,
+                "head_sha": head_sha,
+                "inspected_git_root": self.worktree,
+                "base_branch": "master",
+            },
+        ), patch(
+            "mcp_server.issue_duplicate_context_fetcher",
+            side_effect=lambda h, o, r, auth, issue_number: (
+                list(open_prs), list(branch_names), {"status": "not_claimed"}
+            ),
+        ):
+            with patch.dict(os.environ, env, clear=True):
+                os.environ["GITEA_ISSUE_LOCK_DIR"] = self.lock_dir.name
+                return mcp_server.gitea_lock_issue(
+                    issue_number=ISSUE,
+                    branch_name=BRANCH,
+                    remote="prgs",
+                    worktree_path=self.worktree,
+                )
+
+
+class TestRecoveryWithOwningPrSucceeds(LockIssueEndToEndBase):
+    def test_dead_session_recovery_with_owning_pr_relocks(self):
+        self.write_durable_lock(pid=dead_pid())
+        result = self.run_lock(open_prs=[owning_pr()])
+
+        self.assertTrue(result["success"])
+        self.assertEqual(result["issue_number"], ISSUE)
+        self.assertEqual(result["branch_name"], BRANCH)
+        self.assertTrue(result["lock_freshness"]["live"])
+        self.assertTrue(result["lock_freshness"]["pid_alive"])
+
+    def test_recovered_lock_records_truthful_provenance(self):
+        prior = dead_pid()
+        self.write_durable_lock(pid=prior)
+        self.run_lock(open_prs=[owning_pr()])
+
+        lock = issue_lock_store.load_issue_lock(
+            remote="prgs",
+            org=ORG,
+            repo=REPO,
+            issue_number=ISSUE,
+            lock_dir=self.lock_dir.name,
+        )
+        record = lock.get("dead_session_recovery") or {}
+        self.assertTrue(record.get("recovered"))
+        self.assertEqual(record.get("prior_session_pid"), prior)
+        self.assertEqual(record.get("replacement_session_pid"), os.getpid())
+        self.assertFalse(record.get("prior_pid_alive"))
+        self.assertEqual(record.get("pr_number"), OWNING_PR)
+        self.assertEqual(record.get("branch_name"), BRANCH)
+        self.assertEqual(record.get("identity"), IDENTITY)
+
+    def test_recovered_lock_is_live_and_proves_pr_ownership(self):
+        """AC6: the persisted lock satisfies update-by-merge's ownership prover."""
+        self.write_durable_lock(pid=dead_pid())
+        self.run_lock(open_prs=[owning_pr()])
+
+        lock = issue_lock_store.load_issue_lock(
+            remote="prgs",
+            org=ORG,
+            repo=REPO,
+            issue_number=ISSUE,
+            lock_dir=self.lock_dir.name,
+        )
+        self.assertTrue(issue_lock_store.is_lease_live(lock))
+
+        env = shared_mutation_env(
+            "test-author-prgs",
+            include_example_repo=True,
+            GITEA_ISSUE_LOCK_DIR=self.lock_dir.name,
+        )
+        with patch.dict(os.environ, env, clear=True):
+            os.environ["GITEA_ISSUE_LOCK_DIR"] = self.lock_dir.name
+            ownership = mcp_server._prove_author_ownership_for_pr(
+                pr_number=OWNING_PR,
+                pr_title=f"fix: something (Closes #{ISSUE})",
+                pr_body=f"Closes #{ISSUE}.",
+                source_branch=BRANCH,
+                remote="prgs",
+                host="gitea.prgs.cc",
+                org=ORG,
+                repo=REPO,
+                worktree_path=self.worktree,
+            )
+        self.assertTrue(ownership["proven"], ownership["reasons"])
+        self.assertTrue(ownership["has_author_lock"])
+        self.assertEqual(ownership["matched_issue"], ISSUE)
+
+
+class TestRecoveryRejectionsEndToEnd(LockIssueEndToEndBase):
+    def assert_lock_refused(self, **kwargs):
+        with self.assertRaises((ValueError, RuntimeError)) as ctx:
+            self.run_lock(**kwargs)
+        return str(ctx.exception)
+
+    def test_competing_pr_still_blocked(self):
+        self.write_durable_lock(pid=dead_pid())
+        message = self.assert_lock_refused(
+            open_prs=[owning_pr(number=OTHER_PR, ref=OTHER_BRANCH)],
+            branch_names=[BRANCH],
+        )
+        self.assertIn("already covers issue", message)
+
+    def test_multiple_linked_open_prs_blocked(self):
+        self.write_durable_lock(pid=dead_pid())
+        message = self.assert_lock_refused(
+            open_prs=[owning_pr(), owning_pr(number=OTHER_PR, ref=OTHER_BRANCH)],
+        )
+        self.assertIn("already covers issue", message)
+
+    def test_owning_pr_on_a_different_head_blocked(self):
+        self.write_durable_lock(pid=dead_pid())
+        self.assert_lock_refused(open_prs=[owning_pr(sha=OTHER_HEAD)])
+
+    def test_lock_registered_to_a_different_worktree_blocked(self):
+        self.write_durable_lock(
+            pid=dead_pid(), worktree=os.path.join(self.worktree, "elsewhere")
+        )
+        self.assert_lock_refused(open_prs=[owning_pr()])
+
+    def test_dirty_worktree_blocked(self):
+        self.write_durable_lock(pid=dead_pid())
+        self.assert_lock_refused(
+            open_prs=[owning_pr()], porcelain=" M gitea_mcp_server.py"
+        )
+
+    def test_worktree_parked_on_another_branch_blocked(self):
+        self.write_durable_lock(pid=dead_pid())
+        self.assert_lock_refused(
+            open_prs=[owning_pr()], current_branch="master"
+        )
+
+    def test_local_head_differing_from_remote_blocked(self):
+        self.write_durable_lock(pid=dead_pid())
+        self.assert_lock_refused(
+            open_prs=[owning_pr()], head_sha=OTHER_HEAD
+        )
+
+    def test_live_prior_pid_blocked(self):
+        self.write_durable_lock(pid=os.getpid())
+        self.assert_lock_refused(open_prs=[owning_pr()])
+
+    def test_new_claim_without_prior_lock_still_requires_base_equivalence(self):
+        """AC10: no durable lock → no recovery → base-equivalence still rules."""
+        self.assert_lock_refused(open_prs=[], branch_names=[])
+
+
+if __name__ == "__main__":
+    unittest.main()

tests/test_post_merge_moot_lease.py

@@ -19,6 +19,8 @@ from unittest.mock import patch
 
 sys.path.insert(0, str(__import__("pathlib").Path(__file__).resolve().parent.parent))
 
+import mcp_server  # noqa: E402
+import post_merge_moot_lease_gate as moot_gate  # noqa: E402
 import reviewer_pr_lease as leases  # noqa: E402
 from mcp_server import (  # noqa: E402
     gitea_acquire_reviewer_pr_lease,
@@ -30,6 +32,13 @@ MERGER_ENV = {
     "GITEA_PROFILE_NAME": "prgs-merger",
     "GITEA_ALLOWED_OPERATIONS": "gitea.read,gitea.pr.comment",
 }
+# #745: applying the terminal marker is reconciler-owned.
+RECONCILER_ENV = {
+    "GITEA_PROFILE_NAME": "prgs-reconciler",
+    "GITEA_ALLOWED_OPERATIONS": "gitea.read,gitea.pr.comment,gitea.pr.close",
+}
+CLEANUP_TASK = moot_gate.CLEANUP_TASK
+REPO_SLUG = "Scaled-Tech-Consulting/Gitea-Tools"
 PR = 487
 ISSUE = 485
 SESSION = "97274-676d20a825c4"
@@ -204,6 +213,8 @@ class TestAcquireToolRefusesMergedPR(unittest.TestCase):
 class TestCleanupTool(unittest.TestCase):
     def setUp(self):
         leases.clear_session_lease()
+        moot_gate._reset_for_testing()
+        self.addCleanup(moot_gate._reset_for_testing)
 
     @patch("mcp_server.get_auth_header", return_value=FAKE_AUTH)
     @patch("mcp_server.api_request")
@@ -223,23 +234,53 @@ class TestCleanupTool(unittest.TestCase):
         self.assertEqual(calls["post"], [])
 
     @patch("mcp_server.verify_preflight_purity", return_value=None)
+    @patch("mcp_server._repository_binding_block", return_value=None)
+    @patch("mcp_server._bound_repository_slug", return_value=REPO_SLUG)
     @patch("mcp_server.get_auth_header", return_value=FAKE_AUTH)
     @patch("mcp_server.api_request")
     def test_apply_posts_released_marker_on_merged_pr(
-            self, mock_api, _auth, _purity):
+            self, mock_api, _auth, _slug, _binding, _purity):
+        """#745: apply is reconciler-only and needs a matching dry run first."""
         side, calls = _api_side_effect(
             pr_state="closed", pr_merged=True, comments=[_lease_comment()])
         mock_api.side_effect = side
-        with patch.dict(os.environ, MERGER_ENV, clear=True):
+        with patch.object(mcp_server, "_preflight_resolved_task",
+                          CLEANUP_TASK), \
+                patch.dict(os.environ, RECONCILER_ENV, clear=True):
+            gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=False, remote="prgs")
             result = gitea_cleanup_post_merge_moot_lease(
                 pr_number=PR, apply=True, remote="prgs")
-        self.assertTrue(result["success"])
+        self.assertTrue(result["success"], result.get("reasons"))
         self.assertTrue(result["cleanup_performed"])
         self.assertEqual(result["released_comment_id"], 9999)
         self.assertEqual(len(calls["post"]), 1)
         self.assertIn("phase: released", calls["post"][0]["payload"]["body"])
         self.assertIn("post-merge-moot", calls["post"][0]["payload"]["body"])
 
+    @patch("mcp_server.verify_preflight_purity", return_value=None)
+    @patch("mcp_server._repository_binding_block", return_value=None)
+    @patch("mcp_server._bound_repository_slug", return_value=REPO_SLUG)
+    @patch("mcp_server.get_auth_header", return_value=FAKE_AUTH)
+    @patch("mcp_server.api_request")
+    def test_merger_can_no_longer_apply(
+            self, mock_api, _auth, _slug, _binding, _purity):
+        """#745: holding gitea.pr.comment is no longer sufficient to apply."""
+        side, calls = _api_side_effect(
+            pr_state="closed", pr_merged=True, comments=[_lease_comment()])
+        mock_api.side_effect = side
+        with patch.object(mcp_server, "_preflight_resolved_task",
+                          CLEANUP_TASK), \
+                patch.dict(os.environ, MERGER_ENV, clear=True):
+            gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=False, remote="prgs")
+            result = gitea_cleanup_post_merge_moot_lease(
+                pr_number=PR, apply=True, remote="prgs")
+        self.assertFalse(result["success"])
+        self.assertFalse(result["cleanup_performed"])
+        self.assertEqual(result["blocker_kind"], "wrong_role")
+        self.assertEqual(calls["post"], [])
+
     @patch("mcp_server.verify_preflight_purity", return_value=None)
     @patch("mcp_server.get_auth_header", return_value=FAKE_AUTH)
     @patch("mcp_server.api_request")

workflow_scope_guard.py

@@ -278,12 +278,19 @@ def assess_root_source_mutation(
     current_branch: str | None = None,
     locked_issue_number: int | None = None,
     role_kind: str | None = None,
+    mutation_task: str | None = None,
 ) -> dict[str, Any]:
     """Fail closed for diagnostic/source edits on the control/root checkout.
 
     Allowed only when the active workspace is under ``branches/``. Dirty
     tracked source/test files on the control checkout always block, including
     temporary/diagnostic/test-only intent.
+
+    #749: ``create_issue`` is a pure remote mutation with no local tree write.
+    When *mutation_task* is create_issue and the control checkout has no dirty
+    source/test files, the missing-worktree signal is suppressed so the
+    sanctioned bootstrap path can proceed. Dirty roots and every other task
+    still fail closed.
     """
     role = (role_kind or "").strip().lower()
     if role == "reconciler":
@@ -302,6 +309,7 @@ def assess_root_source_mutation(
     dirty_src = dirty_source_files(porcelain_status)
     reasons: list[str] = []
     blocker_kind: str | None = None
+    create_issue_bootstrap = False
 
     if not under_branches and workspace == root and dirty_src:
         # Root workspace with source dirtiness is unattributed root WIP.
@@ -313,32 +321,50 @@ def assess_root_source_mutation(
         )
         blocker_kind = BLOCKER_ROOT_DIAGNOSTIC_EDIT
 
+    # Lazy import keeps workflow_scope_guard free of circular import at module load.
+    try:
+        import create_issue_bootstrap as _cib
+    except Exception:  # pragma: no cover - import always available in-tree
+        _cib = None
+
     if (
         not under_branches
         and workspace == root
         and not dirty_src
         and role == "author"
     ):
-        # Explicit missing-worktree signal for force-on author entrypoints.
+        if _cib is not None and _cib.is_create_issue_task(mutation_task):
-        reasons.append(
+            # #749: clean-root create_issue is the sanctioned bootstrap path.
-            "author source/test mutation from the stable control checkout is "
+            create_issue_bootstrap = True
-            "forbidden; bind an issue-backed worktree under branches/ first"
+        else:
-        )
+            # Explicit missing-worktree signal for force-on author entrypoints.
-        blocker_kind = BLOCKER_MISSING_WORKTREE
+            reasons.append(
+                "author source/test mutation from the stable control checkout is "
+                "forbidden; bind an issue-backed worktree under branches/ first"
+            )
+            blocker_kind = BLOCKER_MISSING_WORKTREE
 
     if reasons:
         kind = blocker_kind or BLOCKER_ROOT_DIAGNOSTIC_EDIT
+        next_action = _NEXT_ACTIONS[kind]
+        if (
+            kind == BLOCKER_ROOT_DIAGNOSTIC_EDIT
+            and _cib is not None
+            and _cib.is_create_issue_task(mutation_task)
+        ):
+            next_action = _cib.EXACT_NEXT_ACTION_BOOTSTRAP
         return {
             "proven": False,
             "block": True,
             "blocker_kind": kind,
-            "exact_next_action": _NEXT_ACTIONS[kind],
+            "exact_next_action": next_action,
             "reasons": reasons,
             "dirty_source_files": dirty_src,
             "workspace_path": workspace,
             "canonical_repo_root": root,
             "under_branches": under_branches,
             "locked_issue_number": locked_issue_number,
+            "create_issue_bootstrap": False,
         }
     return {
         "proven": True,
@@ -351,6 +377,7 @@ def assess_root_source_mutation(
         "canonical_repo_root": root,
         "under_branches": under_branches,
         "locked_issue_number": locked_issue_number,
+        "create_issue_bootstrap": create_issue_bootstrap,
     }
 
 
@@ -365,6 +392,7 @@ def assess_production_mutation_guards(
     role_kind: str | None = None,
     require_author_lock: bool = False,
     in_test_mode: bool = False,
+    mutation_task: str | None = None,
 ) -> dict[str, Any]:
     """Compose root + scope production guards when they must be active (#683)."""
     if not production_guards_active(in_test_mode=in_test_mode):
@@ -385,6 +413,7 @@ def assess_production_mutation_guards(
         current_branch=current_branch,
         locked_issue_number=locked_issue_number,
         role_kind=role_kind,
+        mutation_task=mutation_task,
     )
     if root_assess["block"]:
         return {**root_assess, "skipped": False}
@@ -408,6 +437,7 @@ def assess_production_mutation_guards(
         "skipped": False,
         "root": root_assess,
         "scope": scope_assess,
+        "create_issue_bootstrap": bool(root_assess.get("create_issue_bootstrap")),
     }