chore(main): release 7.181.2

[github-actions]

PR-review consolidated micro prompt ordering

What changed

PR-review base and micro lanes with explicit owned_workflow_lanes now apply the controller-owned explorer output contract before the PR workflow contract. Operator-authored [CANDIDATE] markers remain rejected with the existing diagnostic.

Why

The PR workflow contract names the owned obligations using controller-generated [CANDIDATE] and [CLEAN] markers. Applying the explorer validator afterward incorrectly classified those controller markers as operator prompt content and blocked valid consolidated micro dispatches.

Migration steps

None.

Breaking changes

None.

Known caveats

Ordinary workflow_lane-only lanes retain the existing contract-first ordering.

---

Stop test_runner's dead-end scope advice and score TESTED SKIPPED as retryable, not failed

Issue: #2756

What changed

• test_runner remediation text (src/tools/test-runner.ts): the guard that rejects scope:"convention"/"graph"/"impact" without files/targets no longer recommends scope:"all" — the exact scope its sibling guard blocks for agent use (env-gated SWARM_ALLOW_FULL_SUITE) and the test_engineer prompt prohibits. The message field now directs the caller to pass a non-empty files array (or targets for framework-native names) with a concrete example, and both response fields (message and error) enumerate every scope the guard covers. Both guards' rejection semantics are unchanged.
• Foreground Stage B verdict settlement (src/hooks/delegation-gate.ts): a [TESTED] | task-N | SKIPPED | ... structured verdict (tests were NOT run — prohibited scope, framework detection none, missing test file) no longer emits stage_b_failed. The task stays in its Stage B eligible state (reviewer_run/pre_check_passed), the reviewer's APPROVED gate proof is preserved, and the stageBCompletion entry is untouched, so the architect can re-dispatch the test gate instead of forcing a coder rework of correct code. A warn log records the skip for the orchestrator. Genuine FAIL verdicts and REVIEWED rejections keep the existing stage_b_failed → rework_required semantics.
• Background Stage B ingestion (src/background/stage-b-gates.ts): structuredStageBVerdict now returns 'skip' for TESTED SKIPPED; ingestBackgroundStageBCompletion consumes the record with the new skipped: true result flag and fires no transition and no proof clearing. StageBIngestionResult gains the optional skipped field.
• Background advisory (src/background/completion-observer.ts): a skipped ingestion publishes skipped (tests not run) — re-dispatch the test gate; reviewer proof preserved; task remains Stage B eligible instead of the generic ingestion failed, so operators can distinguish a retryable skip from a hard failure.

Why

An agent that followed the tool's own advice (scope:"all") could not succeed — the recommended scope is blocked — and models without a natural files: habit (observed: Kimi K2.7 Code, 10–11 identical calls) looped until the repetition breaker fired. The prompt's SKIP CONDITION 1 then legitimately produced a [TESTED] ... SKIPPED verdict, which the gate scored as a code failure: rework_required plus deletion of the reviewer's approval for code that was correct and passing (python -m pytest green). Together with #2755 (no autonomous exit from rework_required, fixed by PR #2760's audited recovery tool), a single tool-argument mistake stranded tasks that only a human could free. This fix removes the wrongful entry: tests-not-run is retryable state, not failure.

Tests

• tests/unit/tools/test-runner-scope-advice.test.ts — guard-2 response fields direct to files/targets, enumerate all three guarded scopes, and never recommend the blocked scope; guard-1 block characterized as unchanged.
• tests/unit/hooks/delegation-gate-stage-b-skipped.test.ts — SKIPPED leaves state reviewer_run with durable reviewer proof and the reviewer completion entry intact; FAIL and REVIEWED REJECTED still go rework_required with proof cleared.
• tests/unit/background/stage-b-gates-skipped-verdict.test.ts — SKIPPED ingest returns skipped: true, no state mutation, proof preserved; FAIL and unparseable output keep the fail-closed rejection.

---

Restart preserves tightened runtime policy and reconciles interrupted work

What changed

• Session-scoped ratchet-tighter QA gate overrides (/swarm qa-gates override) are now durable runtime policy: a new project-DB table
  (qa_gate_session_override, full durability class) is written
  durable-first by the command and restored by rehydrateState, so a host
  restart preserves the effective tightened gates instead of silently
  reverting to the spec-level profile. The override row is deleted in
  lockstep with the session (explicit end, 2-hour stale sweep,
  /swarm reset-session), and a session with no tightened gates keeps no
  row.
• The get_qa_gate_profile tool now also reports the calling session's
  overrides and the effective gates (profile merged with those
  overrides), so an agent can inspect effective runtime policy after a
  restart through the registered surface.
• An execution interrupted by a process boundary (serialized
  delegationActive: true) now reconciles to a bounded, owner-named
  outcome instead of expiring silently: a durable artifact
  (.swarm/session/restart-reconciliation.json, 50-entry FIFO deduped by
  session+task) plus a one-shot advisory on the restored session, both
  naming the owning session, agent, and task and classifying the outcome as
  interrupted/UNKNOWN — absence is never treated as success. Ephemeral
  authority still expires exactly as before; only the silence was a defect.
• New operator documentation (docs/restart-reconciliation.md): the
  durable-vs-ephemeral field classification for the restart boundary, the
  operator-visible reconciliation states, the restart/inspect runbook, and
  the human-resolution rule for external effects.

Why

Issue #2668 (Workstream D slot D15): restart needed to preserve durable
workflow and QA policy while discarding ephemeral execution authority, and
the previously silent expiry of interrupted work left "was interrupted
mid-execution" indistinguishable from "was idle" — an operator could not
tell uncertain from clean, and a valid tightened policy was lost on every
restart.

Tests

• tests/unit/db/qa-gate-session-override.test.ts — service contract:
  ratchet-only merge, fail-closed malformed rows, clear idempotency,
  empty-overrides invariant, absent-DB read-only behavior.
• tests/unit/session/restart-policy-reconciliation.test.ts — boundary
  contract: override survives rehydrate (and a double restart), new sessions
  start clean, fail-open restore, owner-named reconciliation artifact +
  advisory, dedupe, bounded cap, both teardown paths delete the durable row.
• tests/unit/tools/get-qa-gate-profile-effective.test.ts — effective-gates
  tool surface (executor + registered schema fields).
• tests/unit/execute-journey/j08-restart-policy.test.ts — registered-host
  journey: boot A tightens through the real command and dies mid-execution;
  boot B (fresh server() over the same durable artifacts) preserves the
  tightened effective gates via get_qa_gate_profile and records the
  owner-named reconciliation.

🤖 I have created a release beep boop

7.181.2 (2026-09-14)

Bug Fixes

• pr-review: preserve controller-owned micro lane markers (af54a41)
• session: mirror the mutable-resetValue clone into the legacy global rehydrate path (9ad3001)
• session: preserve durable QA policy and owner-named reconciliation across restart (744a241)
• session: prune orphaned override rows and repair advisory dedupe (#2668 review) (5330da1)
• workflow: retryable TESTED SKIPPED verdicts and non-dead-end test_runner scope advice (#2756) (9462f32)

---

This PR was generated with Release Please. See documentation.

[github-actions]

Drift check report

Found 2 drift finding(s): 0 error, 0 warning, 2 notice.

required-check-contract (2)

• 🔵 notice scripts/required-check-contract.json: [RULESET_DIVERGENCE] intended-required context "drift" is not yet required by the captured ruleset
• 🔵 notice scripts/required-check-contract.json: [RULESET_DIVERGENCE] intended-required context "drift" is not present for every expected event in captured external workflow evidence

[github-actions]

🤖 Created releases:

• v7.181.2

🌻