From 5b8c236f3e757f712d65a58bb00839d5f8eb5130 Mon Sep 17 00:00:00 2001
From: SteveBot <1153461+unbraind@users.noreply.github.com>
Date: Fri, 16 Oct 2026 19:04:47 +0200
Subject: [PATCH 1/3] fix(closeout): refuse unverified required-check policies

---
 .agents/pm/history/pm-github-f1ev.jsonl  | 22 ++++++++
 .agents/pm/issues/pm-github-f1ev.toon    | 35 ++++++++++++
 .github/workflows/pm-github-closeout.yml | 16 ++++--
 README.md                                | 42 ++++++++++-----
 closeout.ts                              | 37 ++++++++-----
 docs/closeout-workflow-caller.yml        | 12 +++--
 index.ts                                 |  1 +
 test/closeout-workflow.test.ts           | 38 ++++++++++++-
 test/closeout.test.ts                    | 68 ++++++++++++++++++------
 9 files changed, 220 insertions(+), 51 deletions(-)
 create mode 100644 .agents/pm/history/pm-github-f1ev.jsonl
 create mode 100644 .agents/pm/issues/pm-github-f1ev.toon

diff --git a/.agents/pm/history/pm-github-f1ev.jsonl b/.agents/pm/history/pm-github-f1ev.jsonl
new file mode 100644
index 00000000..520b4745
--- /dev/null
+++ b/.agents/pm/history/pm-github-f1ev.jsonl
@@ -0,0 +1,22 @@
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:04:00.890Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"create","patch":[{"op":"add","path":"/metadata/id","value":"pm-github-f1ev"},{"op":"add","path":"/metadata/title","value":"Refuse closeout when required-check metadata is unverified"},{"op":"add","path":"/metadata/description","value":"Fix https://github.com/unbraind/pm-github/issues/138. Require complete protection metadata or explicit caller-declared contexts; harden caller permissions."},{"op":"add","path":"/metadata/type","value":"Issue"},{"op":"add","path":"/metadata/status","value":"open"},{"op":"add","path":"/metadata/priority","value":1},{"op":"add","path":"/metadata/tags","value":["closeout","validation"]},{"op":"add","path":"/metadata/created_at","value":"2026-10-16T16:04:00.890Z"},{"op":"add","path":"/metadata/updated_at","value":"2026-10-16T16:04:00.890Z"},{"op":"add","path":"/metadata/author","value":"codex-agent"},{"op":"add","path":"/metadata/acceptance_criteria","value":"Unavailable metadata without declarations never closes or publishes; declared contexts require success; metadata-backed behavior remains unchanged; npm and Bun release gates pass."}],"before_hash":"3cc22dff72be7b14824654a7a64ea62b04799939b2fee54c1b5f52ca60bf6df0","after_hash":"8a59cdf7c8f5aa748ad4d926036cf93b490dedc10cbf453b93effa548dc28929","item_hash_version":3,"message":"","context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"7fe69ec357b27d952e3a8de21050066fb7a029cdabb6e6f16012a56da48a45fe"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:04:10.258Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"claim","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:04:10.258Z"},{"op":"add","path":"/metadata/assignee","value":"codex-agent"},{"op":"add","path":"/metadata/claim_principal","value":"codex-agent"}],"before_hash":"8a59cdf7c8f5aa748ad4d926036cf93b490dedc10cbf453b93effa548dc28929","after_hash":"6980e4944bedadfec9082d3d7a3b873dfe11522b019dfd491d9beb45a9e12626","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"5d4d41f1e0c264dec94360a66c9d3d84e63e1af6467c9182bf15e9d9b1bf419b"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:04:12.364Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"update","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:04:12.364Z"},{"op":"replace","path":"/metadata/status","value":"in_progress"}],"before_hash":"6980e4944bedadfec9082d3d7a3b873dfe11522b019dfd491d9beb45a9e12626","after_hash":"2e9d29b5341200b1eda7f7c964c6ba92df21685dc117fc0c7401ea869d0c35d6","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"5b01f5e28064d3f963a9f91666588fad9fd2e5ed0d7d2b71d4870e29716f8fcd"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:04:15.081Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:04:15.081Z"},{"op":"add","path":"/metadata/comments","value":[{"created_at":"2026-10-16T16:04:15.081Z","author":"codex-agent","text":"Plan: distinguish unavailable protection metadata from a readable empty policy. Refuse with required_checks_unverified unless required-checks declares contexts; retain readable app-bound requirements. Exercise real command and workflow shell against injected HTTP evidence and real trackers. Keep workflow edits minimal around open PR #139."}]}],"before_hash":"2e9d29b5341200b1eda7f7c964c6ba92df21685dc117fc0c7401ea869d0c35d6","after_hash":"235d98302d306b4446e87b545d0cf9e9d3f83cf6df49fa384d4bb5bd0461cd65","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"f8e4b7a50cd7e98fdc79c23c47ab67a821e373f3152311d83fd5bafbcaa380f6"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:06:39.639Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"files_add","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:06:39.639Z"},{"op":"add","path":"/metadata/files","value":[{"path":".github/workflows/pm-github-closeout.yml","scope":"project"},{"path":"closeout.ts","scope":"project"},{"path":"index.ts","scope":"project"},{"path":"test/closeout-workflow.test.ts","scope":"project"},{"path":"test/closeout.test.ts","scope":"project"}]}],"before_hash":"235d98302d306b4446e87b545d0cf9e9d3f83cf6df49fa384d4bb5bd0461cd65","after_hash":"99cb9bf8723f9be466c70da32b44d5b366b8c68d6e18d4d2027de3d2b100e6a9","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"e47b8becbe313a63896142a6f5bccc8c458917fcd2fd3c754fe0a3fff4b847e1"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:06:45.909Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"docs_add","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:06:45.909Z"},{"op":"add","path":"/metadata/docs","value":[{"path":"docs/closeout-workflow-caller.yml","scope":"project"},{"path":"README.md","scope":"project"}]}],"before_hash":"99cb9bf8723f9be466c70da32b44d5b366b8c68d6e18d4d2027de3d2b100e6a9","after_hash":"4839048ee250e05ba87aaeccc0f7c515538d10f4c9b7c8dc814de470aa73eb5c","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"eb5ff08804692cdb8cec05a7c5c3b614474c7c8ae36147ffe47ad26cc9c98d3c"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:06:50.000Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"tests_add","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:06:50.000Z"},{"op":"add","path":"/metadata/tests","value":[{"command":"npm run build && npm run build:test && node --test test/closeout.test.ts test/closeout-workflow.test.ts","scope":"project","timeout_seconds":1800,"provenance":{"author":"codex-agent","created_at":"2026-10-16T16:06:49.922Z","source_kind":"local_mutation","source_ref":"fix/closeout-required-checks-unverified"}}]}],"before_hash":"4839048ee250e05ba87aaeccc0f7c515538d10f4c9b7c8dc814de470aa73eb5c","after_hash":"1bcad5aedc7a47d71211ab0e5fac2728b705ff9a5e8e65547adb75c030da8e3e","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"5b31f186680470ea2d4f66af24e4a2219cc3874bbbdb408fa8c3ba2542fd9e99"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:08:21.292Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/1","value":{"created_at":"2026-10-16T16:08:21.292Z","author":"codex-agent","text":"Implementation distinguishes metadata gaps from a readable empty policy. required_checks_unverified blocks workflow publication; required-checks verifies declared contexts plus readable app-bound requirements. Regression coverage uses actual workflow bash, a packed extension, injected HTTP APIs, real SDK tracker mutations and an isolated Git remote. Caller write scopes now belong only to gated-closeout."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:08:21.292Z"}],"before_hash":"1bcad5aedc7a47d71211ab0e5fac2728b705ff9a5e8e65547adb75c030da8e3e","after_hash":"ec44bc6423b950a7bded523297563684d527d62813c5ec39110b31007c16f359","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"2c9a9646564c03c1786159a2c975c677b003e5444621e33e632c65b18896d9cc"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:10:06.013Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/2","value":{"created_at":"2026-10-16T16:10:06.013Z","author":"codex-agent","text":"Behavioural revert proof: restored only checkCloseoutHead to its previous implementation (retaining the added argument for compilation), rebuilt dist, and ran node --test --test-name-pattern='required-check metadata boundary (undeclared|missing) ' test/closeout-workflow.test.ts. Both tests failed: each observed closed=1 where closed=0 was required (0 passed, 2 failed). Fixed function restored and rebuilt. This exercises the packed command and real workflow rather than SDK mocks."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:10:06.013Z"}],"before_hash":"ec44bc6423b950a7bded523297563684d527d62813c5ec39110b31007c16f359","after_hash":"2aa864abc32438faae1ee0908dfbfb2f47baf1c4261888b0f319041a0c40826d","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"c089383fe3ca6847fdd37f88883cc62d063e5588d04dd2fd9e19e9b035696065"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:10:32.730Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"tests_remove","patch":[{"op":"remove","path":"/metadata/tests"},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:10:32.730Z"}],"before_hash":"2aa864abc32438faae1ee0908dfbfb2f47baf1c4261888b0f319041a0c40826d","after_hash":"57ed5b667cbb2b9974ccaef871e887ed669a9f2acb1d6637395983ef84fc173e","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"5778a77b9ec1b516d04171cbb701eaecab84fb70655fcf195f018d8f4c8f2c60"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:10:43.136Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"tests_add","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:10:43.136Z"},{"op":"add","path":"/metadata/tests","value":[{"command":"npm run build && npm run build:test && node --test --test-name-pattern=\"required-check metadata boundary|declared contexts|declared names\" test/closeout.test.ts test/closeout-workflow.test.ts","scope":"project","timeout_seconds":1800,"provenance":{"author":"codex-agent","created_at":"2026-10-16T16:10:42.723Z","source_kind":"local_mutation","source_ref":"fix/closeout-required-checks-unverified"}}]}],"before_hash":"57ed5b667cbb2b9974ccaef871e887ed669a9f2acb1d6637395983ef84fc173e","after_hash":"2af150774079ecb6bac9a74078816899b06324a48e08e55cf181d5e65636b4ee","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"6b5628fb40967071c4585cae635231544d5a386fe53605777b685b094e4bd511"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:11:08.009Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/3","value":{"created_at":"2026-10-16T16:11:08.008Z","author":"codex-agent","text":"Workflow revert proof: removed only required_checks_unverified from the real workflow refusal guard and ran node --test --test-name-pattern='required-check metadata boundary undeclared ' test/closeout-workflow.test.ts. The regression failed (0 passed, 1 failed): workflow exit was 0 instead of required refusal exit 1. Guard restored. Head-function revert separately proved unsafe closures for both absent and missing declarations."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:11:08.009Z"}],"before_hash":"2af150774079ecb6bac9a74078816899b06324a48e08e55cf181d5e65636b4ee","after_hash":"6594c78339011416a76b0f02f5f823904d25f402788a334c45eb67057f2ea7b6","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"efba3c7e4e710044dd51be1a9a4d21ea347898d95f5995b89ed32c4034ac16b0"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:16:08.002Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/4","value":{"created_at":"2026-10-16T16:16:08.002Z","author":"codex-agent","text":"The real workflow regression caught CLI normalization: the host passes requiredChecks while direct core callers use required-checks. Normalize both spellings before validating declarations. The initial linked run reported 4 passed and 5 failed; successful declarations had been ignored. The fix and refusal diagnostic compatibility are now being revalidated before final gates."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:16:08.002Z"}],"before_hash":"6594c78339011416a76b0f02f5f823904d25f402788a334c45eb67057f2ea7b6","after_hash":"85b103f6d93156f58be5327368a30c42f852b3454cff7082083770af1b3a2d23","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"e249a52912115427b0524a1575d09ba967060a58d013a50ae2be852b6cbdecf3"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:20:35.758Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/5","value":{"created_at":"2026-10-16T16:20:35.758Z","author":"codex-agent","text":"Pinned pm linked test run passed 9/9, zero failures, skips, cancellations or todos. Exact command: npm run build && npm run build:test && node --test --test-name-pattern=\"required-check metadata boundary|declared contexts|declared names\" test/closeout.test.ts test/closeout-workflow.test.ts. Covers default refusal/no publication, missing/neutral/skipped/pending declarations, successful declared closure, readable metadata precedence, exact context names, latest legacy statuses, optional failures and preserved app bindings. Both flag spellings now work through the actual command. Earlier pre-fix full runs were stopped and will not be used as gate evidence."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:20:35.758Z"}],"before_hash":"85b103f6d93156f58be5327368a30c42f852b3454cff7082083770af1b3a2d23","after_hash":"e2f5032fdea31a6464512c63a766f22685482aa52c38650da55396ad7f59b6f0","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"806ed075544dc819b99092003c15ee7382afc29e3ceb072a0d8892d4ed6c3897"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T16:45:11.663Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/6","value":{"created_at":"2026-10-16T16:45:11.663Z","author":"codex-agent","text":"Fresh npm run release:check passed in a clean HOME with GIT_CONFIG_GLOBAL=/dev/null and GIT_CONFIG_NOSYSTEM=1: 711/711 tests, zero failures/skips/cancellations/todos. Exact coverage across 13 authored modules: lines and statements 10082/10082, functions 256/256, branches 2858/2858, all four dimensions 100 percent at unchanged thresholds. Typecheck, build, docstrings, lint, zero duplication, privacy, production audit (zero vulnerabilities), package dry-run, changelog and publish-attestation gates passed. Clean-home pm health --strict-exit --require-merge-drivers also passed; normal host health reports an environment-specific extension version skew and existing stale-item warnings. Bun release gate is running next."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T16:45:11.663Z"}],"before_hash":"e2f5032fdea31a6464512c63a766f22685482aa52c38650da55396ad7f59b6f0","after_hash":"4fbb41a04bd6ca35bcd08cb7eb8885709029448b89a30d3db3a0ba01800f3c46","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"90bb64225749e8865bbaddc656afd35a24966a0e4ae0ae52b02c404e1df33922"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T17:04:25.496Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/7","value":{"created_at":"2026-10-16T17:04:25.496Z","author":"codex-agent","text":"bun run release:check passed in the same clean-home/Git identity environment: 711/711 tests, zero failures/skips/cancellations/todos, 13 authored modules with exact 100/100/100/100 coverage (10082 lines/statements, 256 functions, 2858 branches). All remaining release gates passed at unchanged thresholds. Both behavioral revert proofs and the passing 9/9 real-command boundary run are recorded above. Caller declarations must remain aligned with required contexts; no release or downstream adoption is part of this change. Preparing the review PR for issue #138; item remains in_progress and will not be closed."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T17:04:25.496Z"}],"before_hash":"4fbb41a04bd6ca35bcd08cb7eb8885709029448b89a30d3db3a0ba01800f3c46","after_hash":"202a7ffac6d16e8565c250b7e5a47a5610a82df47ee721bc1a8faa2a7a040b21","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"3aa7da4c279dead9146dfad652a5b9ce1ca1c3246d8639c307fd5ed029679e7a"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T17:04:26.562Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"release","patch":[{"op":"remove","path":"/metadata/claim_principal"},{"op":"remove","path":"/metadata/assignee"},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T17:04:26.562Z"}],"before_hash":"202a7ffac6d16e8565c250b7e5a47a5610a82df47ee721bc1a8faa2a7a040b21","after_hash":"5e22ce94e9f1512e2a1bdd25d3878cf2af1f3c6620c8e0d6539d1e2f64bceca6","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"8592b0c0960a595d86481d150d967630ab98a869d998dc0d78950f2db989322e"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T17:05:39.029Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"claim","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T17:05:39.029Z"},{"op":"add","path":"/metadata/assignee","value":"codex-agent"},{"op":"add","path":"/metadata/claim_principal","value":"codex-agent"}],"before_hash":"5e22ce94e9f1512e2a1bdd25d3878cf2af1f3c6620c8e0d6539d1e2f64bceca6","after_hash":"7bed1df316447fac8d5b4aaa778e9e74c518db589c51329d1dfa7a9db1db8383","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"a8d9cd74a58bd449552bbef50cda5e0782f5d60f044926eb0a8508d99e683f2c"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T17:05:41.112Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/8","value":{"created_at":"2026-10-16T17:05:41.112Z","author":"codex-agent","text":"PR #139 merged during validation. Rebased onto main commit 6d9cc23 and resolved the caller-template conflict by preserving the self-versioned release SHA/tag comment plus this change's job-only permissions and required-checks input. No self-versioning implementation was changed. Previous 711/711 npm and Bun receipts predate this integration; rerunning the full gates on the rebased branch before delivery."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T17:05:41.112Z"}],"before_hash":"7bed1df316447fac8d5b4aaa778e9e74c518db589c51329d1dfa7a9db1db8383","after_hash":"f6d98341e2f3e8b7794a72bd2b2a4d964ad957884ffd8cb1455a71649bd8653a","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"6d23d2c15e60ca3bccbad6da62aa4d25458e7652bce95f026b3b509d576c0a89"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T17:41:09.873Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/9","value":{"created_at":"2026-10-16T17:41:09.873Z","author":"codex-agent","text":"Rebased acceptance on main 6d9cc23: npm run release:check passed 747/747 tests with zero failures/skips/cancellations/todos. Exact coverage remains 13 modules, 10082/10082 lines and statements, 256/256 functions, 2858/2858 branches (100/100/100/100; thresholds unchanged). All release gates passed. Dedicated integration checks also passed: node --test test/self-versioned-workflows.test.ts (36/36) and node --test --test-name-pattern='closeout contract' test/closeout-workflow.test.ts (1/1). The rebased Bun gate is now running."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T17:41:09.873Z"}],"before_hash":"f6d98341e2f3e8b7794a72bd2b2a4d964ad957884ffd8cb1455a71649bd8653a","after_hash":"3d91279c7db59105c0a18c8cca829ea0d85e7e78626f392452d01f00515c4a83","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"9b10d8ca06009ddbe49ee6cb9b12827f87b98dabb45e35d9987cd3605bd75527"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T18:06:25.089Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/10","value":{"created_at":"2026-10-16T18:06:25.089Z","author":"codex-agent","text":"Final rebased acceptance: npm run release:check and bun run release:check both passed 747/747 tests, zero failures/skips/cancellations/todos, exact 100/100/100/100 coverage across 13 authored modules (10082 lines/statements, 256 functions, 2858 branches). All release gates passed at unchanged thresholds on the implementation rebased over #139. Behavioral head-function and workflow-guard revert proofs remain recorded above; real-command boundary tests passed 9/9. Ready for review; no release, merge, downstream adoption or item close is performed. Caller-declared context lists must stay aligned with required checks."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T18:06:25.089Z"}],"before_hash":"3d91279c7db59105c0a18c8cca829ea0d85e7e78626f392452d01f00515c4a83","after_hash":"a30a356d83e28ea153b2339786a91336952cfcbade9446f834bee5e98f584fff","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"e81721bfbfc62b985ed775994060faa16260233d84902a1b53e1d9ac11b07325"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T18:06:26.222Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"release","patch":[{"op":"remove","path":"/metadata/claim_principal"},{"op":"remove","path":"/metadata/assignee"},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T18:06:26.222Z"}],"before_hash":"a30a356d83e28ea153b2339786a91336952cfcbade9446f834bee5e98f584fff","after_hash":"95a01ca6cd3e4bc3e6fcaec8f066166e9dd8fc809ad76ff49c536c9b22248393","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"2277a5f3985a79d596dcc9aa2cae1e8d4291368307da1269f08ad01777e82515"}
diff --git a/.agents/pm/issues/pm-github-f1ev.toon b/.agents/pm/issues/pm-github-f1ev.toon
new file mode 100644
index 00000000..15bffe03
--- /dev/null
+++ b/.agents/pm/issues/pm-github-f1ev.toon
@@ -0,0 +1,35 @@
+id: pm-github-f1ev
+title: Refuse closeout when required-check metadata is unverified
+description: "Fix https://github.com/unbraind/pm-github/issues/138. Require complete protection metadata or explicit caller-declared contexts; harden caller permissions."
+type: Issue
+status: in_progress
+priority: 1
+tags[2]: closeout,validation
+created_at: "2026-10-16T16:04:00.890Z"
+updated_at: "2026-10-16T18:06:26.222Z"
+author: codex-agent
+acceptance_criteria: Unavailable metadata without declarations never closes or publishes; declared contexts require success; metadata-backed behavior remains unchanged; npm and Bun release gates pass.
+comments[11]{created_at,author,text}:
+  "2026-10-16T16:04:15.081Z",codex-agent,"Plan: distinguish unavailable protection metadata from a readable empty policy. Refuse with required_checks_unverified unless required-checks declares contexts; retain readable app-bound requirements. Exercise real command and workflow shell against injected HTTP evidence and real trackers. Keep workflow edits minimal around open PR #139."
+  "2026-10-16T16:08:21.292Z",codex-agent,"Implementation distinguishes metadata gaps from a readable empty policy. required_checks_unverified blocks workflow publication; required-checks verifies declared contexts plus readable app-bound requirements. Regression coverage uses actual workflow bash, a packed extension, injected HTTP APIs, real SDK tracker mutations and an isolated Git remote. Caller write scopes now belong only to gated-closeout."
+  "2026-10-16T16:10:06.013Z",codex-agent,"Behavioural revert proof: restored only checkCloseoutHead to its previous implementation (retaining the added argument for compilation), rebuilt dist, and ran node --test --test-name-pattern='required-check metadata boundary (undeclared|missing) ' test/closeout-workflow.test.ts. Both tests failed: each observed closed=1 where closed=0 was required (0 passed, 2 failed). Fixed function restored and rebuilt. This exercises the packed command and real workflow rather than SDK mocks."
+  "2026-10-16T16:11:08.008Z",codex-agent,"Workflow revert proof: removed only required_checks_unverified from the real workflow refusal guard and ran node --test --test-name-pattern='required-check metadata boundary undeclared ' test/closeout-workflow.test.ts. The regression failed (0 passed, 1 failed): workflow exit was 0 instead of required refusal exit 1. Guard restored. Head-function revert separately proved unsafe closures for both absent and missing declarations."
+  "2026-10-16T16:16:08.002Z",codex-agent,"The real workflow regression caught CLI normalization: the host passes requiredChecks while direct core callers use required-checks. Normalize both spellings before validating declarations. The initial linked run reported 4 passed and 5 failed; successful declarations had been ignored. The fix and refusal diagnostic compatibility are now being revalidated before final gates."
+  "2026-10-16T16:20:35.758Z",codex-agent,"Pinned pm linked test run passed 9/9, zero failures, skips, cancellations or todos. Exact command: npm run build && npm run build:test && node --test --test-name-pattern=\"required-check metadata boundary|declared contexts|declared names\" test/closeout.test.ts test/closeout-workflow.test.ts. Covers default refusal/no publication, missing/neutral/skipped/pending declarations, successful declared closure, readable metadata precedence, exact context names, latest legacy statuses, optional failures and preserved app bindings. Both flag spellings now work through the actual command. Earlier pre-fix full runs were stopped and will not be used as gate evidence."
+  "2026-10-16T16:45:11.663Z",codex-agent,"Fresh npm run release:check passed in a clean HOME with GIT_CONFIG_GLOBAL=/dev/null and GIT_CONFIG_NOSYSTEM=1: 711/711 tests, zero failures/skips/cancellations/todos. Exact coverage across 13 authored modules: lines and statements 10082/10082, functions 256/256, branches 2858/2858, all four dimensions 100 percent at unchanged thresholds. Typecheck, build, docstrings, lint, zero duplication, privacy, production audit (zero vulnerabilities), package dry-run, changelog and publish-attestation gates passed. Clean-home pm health --strict-exit --require-merge-drivers also passed; normal host health reports an environment-specific extension version skew and existing stale-item warnings. Bun release gate is running next."
+  "2026-10-16T17:04:25.496Z",codex-agent,"bun run release:check passed in the same clean-home/Git identity environment: 711/711 tests, zero failures/skips/cancellations/todos, 13 authored modules with exact 100/100/100/100 coverage (10082 lines/statements, 256 functions, 2858 branches). All remaining release gates passed at unchanged thresholds. Both behavioral revert proofs and the passing 9/9 real-command boundary run are recorded above. Caller declarations must remain aligned with required contexts; no release or downstream adoption is part of this change. Preparing the review PR for issue #138; item remains in_progress and will not be closed."
+  "2026-10-16T17:05:41.112Z",codex-agent,PR #139 merged during validation. Rebased onto main commit 6d9cc23 and resolved the caller-template conflict by preserving the self-versioned release SHA/tag comment plus this change's job-only permissions and required-checks input. No self-versioning implementation was changed. Previous 711/711 npm and Bun receipts predate this integration; rerunning the full gates on the rebased branch before delivery.
+  "2026-10-16T17:41:09.873Z",codex-agent,"Rebased acceptance on main 6d9cc23: npm run release:check passed 747/747 tests with zero failures/skips/cancellations/todos. Exact coverage remains 13 modules, 10082/10082 lines and statements, 256/256 functions, 2858/2858 branches (100/100/100/100; thresholds unchanged). All release gates passed. Dedicated integration checks also passed: node --test test/self-versioned-workflows.test.ts (36/36) and node --test --test-name-pattern='closeout contract' test/closeout-workflow.test.ts (1/1). The rebased Bun gate is now running."
+  "2026-10-16T18:06:25.089Z",codex-agent,"Final rebased acceptance: npm run release:check and bun run release:check both passed 747/747 tests, zero failures/skips/cancellations/todos, exact 100/100/100/100 coverage across 13 authored modules (10082 lines/statements, 256 functions, 2858 branches). All release gates passed at unchanged thresholds on the implementation rebased over #139. Behavioral head-function and workflow-guard revert proofs remain recorded above; real-command boundary tests passed 9/9. Ready for review; no release, merge, downstream adoption or item close is performed. Caller-declared context lists must stay aligned with required checks."
+files[5]{path,scope}:
+  .github/workflows/pm-github-closeout.yml,project
+  closeout.ts,project
+  index.ts,project
+  test/closeout-workflow.test.ts,project
+  test/closeout.test.ts,project
+tests[1]{command,scope,timeout_seconds,provenance{author,created_at,source_kind,source_ref}}:
+  "npm run build && npm run build:test && node --test --test-name-pattern=\"required-check metadata boundary|declared contexts|declared names\" test/closeout.test.ts test/closeout-workflow.test.ts",project,1800,codex-agent,"2026-10-16T16:10:42.723Z",local_mutation,fix/closeout-required-checks-unverified
+docs[2]{path,scope}:
+  docs/closeout-workflow-caller.yml,project
+  README.md,project
+body: ""
diff --git a/.github/workflows/pm-github-closeout.yml b/.github/workflows/pm-github-closeout.yml
index 0f0c646b..57d33a1d 100644
--- a/.github/workflows/pm-github-closeout.yml
+++ b/.github/workflows/pm-github-closeout.yml
@@ -31,6 +31,11 @@ on:
         type: string
         required: false
         default: "ci.yml"
+      required-checks:
+        description: "Whitespace-separated required contexts to verify when protection metadata is unavailable"
+        type: string
+        required: false
+        default: ""
       node-version:
         description: "Node.js version for the pm CLI"
         type: string
@@ -59,6 +64,7 @@ jobs:
       REQUESTED_PM_GITHUB_VERSION: ${{ inputs.pm-github-version }}
       NPM_PROJECT_DIR: ${{ inputs.npm-project-dir }}
       POST_CLOSEOUT_SCRIPT: ${{ inputs.post-closeout-script }}
+      REQUIRED_CHECKS: ${{ inputs.required-checks }}
       CI_WORKFLOWS: ${{ inputs.ci-workflows }}
       CALLING_REPOSITORY: ${{ github.repository }}
     steps:
@@ -283,7 +289,11 @@ jobs:
           set -euo pipefail
           # Full history catches up after missed schedules; the command skips terminal items.
           # Keep the report private until the proposed change has passed the privacy gate.
-          if ! pm github closeout --repo "${REPOSITORY}" --since 1970-01-01 --apply --json > "${RUNNER_TEMP}/pm-github-closeout.json" 2> "${RUNNER_TEMP}/pm-github-closeout.err"; then
+          required_args=()
+          if [[ -n "${REQUIRED_CHECKS:-}" ]]; then
+            required_args+=(--required-checks "${REQUIRED_CHECKS}")
+          fi
+          if ! pm github closeout --repo "${REPOSITORY}" --since 1970-01-01 --apply --json "${required_args[@]}" > "${RUNNER_TEMP}/pm-github-closeout.json" 2> "${RUNNER_TEMP}/pm-github-closeout.err"; then
             echo "::error::Closeout refused; check token read access to pull requests, checks and branch rules, and tracker close validation. No branch was published."
             exit 1
           fi
@@ -291,8 +301,8 @@ jobs:
           # Only a local tracker refusal makes missing head evidence actionable.
           # Foreign, terminal, excluded and unlinked PRs cannot block this run.
           jq -e 'all(.plan[]; (.code | type) == "string")' "${RUNNER_TEMP}/pm-github-closeout.json" >/dev/null
-          if jq -e 'any(.plan[]; .action == "skip" and .code == "head_evidence_unavailable")' "${RUNNER_TEMP}/pm-github-closeout.json" >/dev/null; then
-            echo "::error::Closeout refused; check token read access to head checks for linked local tracker items. No branch was published."
+          if jq -e 'any(.plan[]; .action == "skip" and (.code == "head_evidence_unavailable" or .code == "required_checks_unverified"))' "${RUNNER_TEMP}/pm-github-closeout.json" >/dev/null; then
+            echo "::error::Closeout refused; check token read access to head checks for linked local tracker items, and declare required-checks when protection metadata is unavailable. No branch was published."
             exit 1
           fi
           # Merge facts are checked before head lookup and have a distinct decision.
diff --git a/README.md b/README.md
index 3d14490e..57f9fbb2 100644
--- a/README.md
+++ b/README.md
@@ -65,14 +65,21 @@ latest status. Check-run, status and ruleset pages are fully read. Rulesets are
 read through GitHub's [effective branch rules API](https://docs.github.com/en/rest/repos/rules#get-rules-for-a-branch).
 The check policy reflects the rules readable at invocation time.
 
-When a protection source is unreadable (HTTP 403/404), or no required contexts
-are configured, the report labels the **fallback** rule. Every readable required
-context still must conclude success, including its app binding. The fallback
-additionally requires at least one head check run, every run completed with
-`success`, `neutral` or `skipped`, and successful latest legacy head statuses.
-It never discards a known requirement because another source is unreadable.
-Missing, malformed or incomplete head evidence, malformed metadata and other
-metadata errors fail closed. The plan reports which rule was used.
+When either protection source is unreadable (HTTP 403/404), the row is ineligible
+with `required_checks_unverified`. Observed green runs cannot prove that every
+required context ran. Supply `--required-checks "CI CodeQL"` to declare a
+whitespace-separated set of required context names when metadata is unavailable.
+Every declared context must conclude **success** on the exact head (or have a
+successful latest legacy status); missing, pending, neutral or skipped contexts
+refuse closure. Readable requirements remain enforced, including app bindings.
+This **declared** policy does not require unrelated optional runs to succeed.
+When both metadata sources are readable, declarations do not replace their policy.
+
+With fully readable metadata and no required contexts configured, the report
+labels the **fallback** rule: at least one head check run must exist, every run
+must complete with `success`, `neutral` or `skipped`, and latest legacy statuses
+must succeed. Missing, malformed or incomplete evidence and other metadata errors
+fail closed. The plan reports which rule was used.
 
 TOON lists actionable `close`/`refuse` rows individually and summarizes routine skips
 as counts by stable code, for example `skipped: {already_terminal: 30,
@@ -105,8 +112,10 @@ structured close fields it supports, without overriding tracker policy:
   otherwise `Delivered by the linked pull request, merged after its head check gate passed.`
 - `actual_result`: `Merged at <time>; required checks satisfied; successful head checks:
   <names>` when branch protection or rulesets name required checks (the list holds every
-  successful head check, optional ones included). Under the fallback rule (no readable
-  required contexts) it reads `Merged at <time>; no readable required checks, so every
+  successful head check, optional ones included). Under the declared policy it reads
+  `Merged at <time>; caller-declared and readable required checks satisfied; successful
+  head checks: <names>`. Under the fallback rule (fully readable, empty policy) it reads
+  `Merged at <time>; no readable required checks, so every
   completed head check run had to succeed, be neutral or be skipped; successful checks:
   <names>`, or `…; no check reported success` when only neutral or skipped runs exist.
   Only the verified merge time, the applied check policy and sorted successful check
@@ -543,6 +552,13 @@ such as `trunk` or `release/stable`. The caller's pinned PM CLI comes from
 inside the workspace and contain `package.json`. The SDK resolves the tracker
 from the repository root. The exact release version resolved from the workflow commit is installed and verified.
 
+`required-checks` is an optional workflow input with an empty default. Set it to
+all required context names (for example `"CI CodeQL"`) when the token cannot read
+branch-protection or ruleset metadata, as is common with `GITHUB_TOKEN`. The workflow
+passes this value as one quoted command argument. Without a declaration, unavailable
+metadata refuses closeout and prevents publishing a branch or PR. Keep this list
+aligned with the caller's required checks; it is a caller-owned policy declaration.
+
 Each run rebuilds the bot branch from the current default branch and invokes
 `pm github closeout --since 1970-01-01 --apply --json`. Scanning all merge history
 avoids losing deliveries during schedule outages; it costs more API reads for
@@ -572,13 +588,15 @@ certify that they passed. Review and required-check enforcement remain with the
 caller's protected branch. A repository concurrency group serializes closeout
 runs, and the explicit lease rejects competing writers.
 
-The caller grants only `contents: write`, `pull-requests: write`, `checks: read`
+The caller template uses top-level `permissions: {}` and grants its
+`gated-closeout` job only `contents: write`, `pull-requests: write`, `checks: read`
 and `actions: write`. Closeout uses `checks: read` instead of sync's `issues: read`
 to retrieve head-check evidence ([GitHub Checks API permissions](https://docs.github.com/en/rest/checks/runs#list-check-runs-for-a-git-reference)).
 The reusable workflow cannot elevate the caller's token
 permissions ([GitHub reusable workflow permissions](https://docs.github.com/en/actions/reference/workflows-and-actions/reusing-workflow-configurations)).
 Repository Actions settings must allow Actions to create PRs. Read denial stops
-closeout before push with a diagnostic, including unavailable head-check evidence
+closeout before push with a diagnostic, including `required_checks_unverified` and
+unavailable head-check evidence
 reported by the command as a skip when it blocks a local, existing, non-terminal,
 non-excluded tracker item. Unlinked PRs and foreign/terminal/excluded links do
 not fail the run for unavailable evidence. Incomplete merge facts use a separate
diff --git a/closeout.ts b/closeout.ts
index af5b3ff6..2682683d 100644
--- a/closeout.ts
+++ b/closeout.ts
@@ -20,13 +20,13 @@ export interface CloseoutRow {
 /** Stable decisions used by automation and compact skip counts. */
 export type CloseoutCode = "already_terminal" | "other_repository" | "duplicate_selection" | "explicit_skip"
   | "unknown_item" | "epic" | "active_children" | "not_delivery" | "close_failed"
-  | "head_checks_passed" | "head_checks_failed" | "head_evidence_unavailable" | "incomplete_merge";
+  | "head_checks_passed" | "head_checks_failed" | "head_evidence_unavailable" | "required_checks_unverified" | "incomplete_merge";
 
 /** Which head-check policy was evaluated for one selected pull request. */
 interface CheckGate {
   pr: number;
   code: CloseoutCode;
-  rule: "required" | "fallback" | "unavailable";
+  rule: "required" | "declared" | "fallback" | "unavailable";
   reason: string;
   passed: string[];
   ok: boolean;
@@ -47,7 +47,7 @@ interface PullRequest {
 /** A required context may be bound to a particular GitHub App. */
 interface RequiredCheck { name: string; app: number | null }
 /** Preserve readable requirements while recording a metadata gap or an empty policy. */
-interface RequiredPolicy { checks: RequiredCheck[]; fallback: boolean }
+interface RequiredPolicy { checks: RequiredCheck[]; fallback: boolean; unavailable: boolean }
 /** Only check-run facts used to decide whether a head passed. */
 interface CheckRun { name: string; head_sha: string; status: string; conclusion: string | null; app: { id: number } }
 /** Legacy commit statuses are newest first; only the latest per context counts. */
@@ -181,15 +181,24 @@ async function requirements(api: CloseoutApi, repo: string, branch: string): Pro
       checks.push({ name: text(check.context), app: typeof app === "number" ? app : null });
     }
   }
-  return { checks, fallback: classic === null || rules === null || checks.length === 0 };
+  return { checks, fallback: checks.length === 0, unavailable: classic === null || rules === null };
 }
 
-/** Verify required checks or the explicitly reported conservative fallback at the PR head. */
-export async function checkCloseoutHead(api: CloseoutApi, repo: string, pr: PullRequest): Promise<CheckGate> {
+/** Verify metadata-backed or caller-declared requirements before considering observed head runs. */
+export async function checkCloseoutHead(api: CloseoutApi, repo: string, pr: PullRequest, declared: string[] = []): Promise<CheckGate> {
   const result: CheckGate = { pr: pr.number, code: "head_evidence_unavailable", rule: "unavailable", reason: "Head check evidence unavailable or incomplete.", passed: [], ok: false };
   try {
     const policy = await requirements(api, repo, pr.base.ref);
-    result.rule = policy.fallback ? "fallback" : "required";
+    if (policy.unavailable) {
+      if (declared.length === 0) {
+        result.code = "required_checks_unverified";
+        result.reason = "Required-check metadata unavailable; declare --required-checks to verify the required contexts.";
+        return result;
+      }
+      policy.checks.push(...declared.map(name => ({ name, app: null })));
+      policy.fallback = false;
+    }
+    result.rule = policy.unavailable ? "declared" : policy.fallback ? "fallback" : "required";
     const runs: CheckRun[] = (await pages(api, `/repos/${repo}/commits/${pr.head.sha}/check-runs?filter=latest&per_page=100`, true)).map(value => {
       const run = record(value);
       const app = record(run.app);
@@ -214,7 +223,8 @@ export async function checkCloseoutHead(api: CloseoutApi, repo: string, pr: Pull
       }
     }
     if (!policy.fallback) {
-      result.reason = "Required head checks concluded success (branch-protection and ruleset rule).";
+      result.reason = policy.unavailable ? "Caller-declared and readable required head checks concluded success."
+        : "Required head checks concluded success (branch-protection and ruleset rule).";
     } else {
       if (runs.length === 0) { result.reason = "Fallback rule: no check runs at head; evidence missing."; return result; }
       if (runs.some(run => run.status !== "completed" || !["success", "neutral", "skipped"].includes(String(run.conclusion)))
@@ -222,7 +232,7 @@ export async function checkCloseoutHead(api: CloseoutApi, repo: string, pr: Pull
         result.reason = "Fallback rule: every head run must be completed with success, neutral or skipped, and head statuses must succeed.";
         return result;
       }
-      result.reason = "Fallback rule: readable required checks succeeded; protection unavailable or no required contexts; every head run completed with success, neutral or skipped; head statuses succeeded.";
+      result.reason = "Fallback rule: protection metadata is readable with no required contexts; every head run completed with success, neutral or skipped; head statuses succeeded.";
     }
     result.passed = [...new Set([...runs.filter(run => run.status === "completed" && run.conclusion === "success").map(run => run.name),
       ...[...statuses.values()].filter(status => status.state === "success").map(status => status.context)])].sort();
@@ -326,12 +336,13 @@ function usage<T>(parse: () => T): T {
 /**
  * Describe the verified check evidence for a passed gate without overstating it.
  *
- * The required rule states that the required checks were satisfied and lists every successful head check (optional
- * ones included). The fallback rule (no readable required contexts)
+ * The required and declared rules state that their required checks were satisfied and list every successful head check (optional
+ * ones included). The fallback rule (fully readable metadata with no required contexts)
  * may pass on neutral or skipped runs alone, so it states that policy and lists successful checks only when any exist.
  */
 export function checkEvidence(gate: CheckGate): string {
   // gate.passed holds every successful head check, optional ones included, so it is never labelled as the required set.
+  if (gate.rule === "declared") return `caller-declared and readable required checks satisfied; successful head checks: ${gate.passed.join(", ")}`;
   if (gate.rule === "required") return `required checks satisfied; successful head checks: ${gate.passed.join(", ")}`;
   const successes = gate.passed.length > 0 ? `successful checks: ${gate.passed.join(", ")}` : "no check reported success";
   return `no readable required checks, so every completed head check run had to succeed, be neutral or be skipped; ${successes}`;
@@ -346,6 +357,8 @@ export async function runCloseout(pmRoot: string, repo: string | undefined, opti
   });
   const since = options.since === undefined ? undefined : usage(() => isoDate(text(options.since)));
   if ((numbers.length === 0) === (since === undefined)) throw new CloseoutUsageError("Select either repeatable --pr <n> or --since <ISO date>.");
+  const declaredOptions = { "required-checks": options["required-checks"] ?? options.requiredChecks };
+  const declared = usage(() => repeated(declaredOptions, "required-checks")).flatMap(value => value.split(/\s+/));
   const skip = new Set(usage(() => repeated(options, "skip")));
   const apply = options.apply === true && options["dry-run"] !== true && options.dryRun !== true;
   if (since !== undefined) {
@@ -376,7 +389,7 @@ export async function runCloseout(pmRoot: string, repo: string | undefined, opti
   for (const number of [...new Set(numbers)].sort((a, b) => a - b)) {
     const pr = pullRequest(await payload(api, `/repos/${repo}/pulls/${number}`), repo, number);
     let gate: CheckGate = { pr: number, code: "incomplete_merge", rule: "unavailable", ok: false, passed: [], reason: "Pull request is not merged or merge evidence is incomplete." };
-    if (mergeEvidence(pr)) gate = await checkCloseoutHead(api, repo, pr);
+    if (mergeEvidence(pr)) gate = await checkCloseoutHead(api, repo, pr, declared);
     checks.push(gate);
     for (const ref of extractCloseoutItems(pr.body ?? "")) {
       let reason: string;
diff --git a/docs/closeout-workflow-caller.yml b/docs/closeout-workflow-caller.yml
index accfc36d..f2eebcf9 100644
--- a/docs/closeout-workflow-caller.yml
+++ b/docs/closeout-workflow-caller.yml
@@ -11,15 +11,17 @@ on:
   schedule:
     - cron: "43 3 * * *"
   workflow_dispatch:
-permissions:
-  contents: write
-  pull-requests: write
-  checks: read
-  actions: write
+permissions: {}
 jobs:
   gated-closeout:
+    permissions:
+      contents: write
+      pull-requests: write
+      checks: read
+      actions: write
     uses: unbraind/pm-github/.github/workflows/pm-github-closeout.yml@RELEASE_COMMIT_SHA # vRELEASE_VERSION
     with:
       npm-project-dir: "." # Use tools/oracle or another contained npm project when needed.
       post-closeout-script: "changelog:full" # Missing script skips; "" disables it.
+      required-checks: "" # Whitespace-separated context names (e.g. "CI CodeQL"); required when protection metadata cannot be read.
       ci-workflows: "ci.yml" # Whitespace-separated workflow filenames; "" disables dispatch.
diff --git a/index.ts b/index.ts
index 819ee3eb..7c8405f7 100644
--- a/index.ts
+++ b/index.ts
@@ -6688,6 +6688,7 @@ export default defineExtension({
         { long: "--repo", value_name: "owner/repo", description: "Repository (or PM_GITHUB_REPO)" },
         { long: "--pr", value_name: "n", repeatable: true, description: "Select a PR number (repeatable; conflicts with --since)" },
         { long: "--since", value_name: "ISO date", description: "Select PRs merged at or after this ISO date/time" },
+        { long: "--required-checks", value_name: "contexts", description: "Whitespace-separated required contexts when protection metadata is unavailable" },
         { long: "--skip", value_name: "id", repeatable: true, description: "Never close this item (repeatable)" },
         { long: "--apply", description: "Close eligible items (default: dry-run)" },
         { long: "--dry-run", description: "Preview only; overrides --apply" },
diff --git a/test/closeout-workflow.test.ts b/test/closeout-workflow.test.ts
index 5733655d..0b8554f3 100644
--- a/test/closeout-workflow.test.ts
+++ b/test/closeout-workflow.test.ts
@@ -146,6 +146,7 @@ esac
   const server = await startMockGithub((req, res) => {
     const url = req.url!;
     if (url.includes("/check-runs")) env.HEAD_LOOKUPS = String(Number(env.HEAD_LOOKUPS ?? 0) + 1);
+    if (env.PROTECTION_DENIED === "true" && (url.includes("/protection/") || url.includes("/rules/branches/"))) { jsonResponse(res, 403, {}); return; }
     if (env.READ_DENIED === "true" || (env.CHECKS_DENIED === "true" && url.includes("/check-runs"))) { jsonResponse(res, 403, { message: "Resource not accessible by integration" }); return; }
     const sha = "a".repeat(40);
     if (url.includes("/pulls?")) jsonResponse(res, 200, env.NO_DELIVERY === "true" ? [] : [{ number: 42, merged_at: "2026-10-15T00:00:00Z", updated_at: "2026-10-15T00:00:00Z" }]);
@@ -154,7 +155,7 @@ esac
       head: { sha }, base: { ref: "trunk", repo: { full_name: "acme/widgets" } } });
     else if (url.includes("/protection/")) jsonResponse(res, 200, { contexts: ["CI"], checks: [] });
     else if (url.includes("/rules/branches/")) jsonResponse(res, 200, []);
-    else if (url.includes("/check-runs")) jsonResponse(res, 200, { total_count: 1, check_runs: [{ name: "CI", head_sha: sha, status: "completed", conclusion: "success", app: { id: 7 } }] });
+    else if (url.includes("/check-runs")) jsonResponse(res, 200, { total_count: 1, check_runs: [{ name: "CI", head_sha: sha, status: env.HEAD_STATUS ?? "completed", conclusion: env.HEAD_CONCLUSION ?? "success", app: { id: 7 } }] });
     else if (url.includes("/statuses")) jsonResponse(res, 200, []);
     else jsonResponse(res, 404, {});
   });
@@ -207,7 +208,10 @@ test("closeout contract keeps sync safety helpers, exact pins, env bindings and
   assert.match(caller, /workflow_dispatch:/);
   assert.match(caller, /npm-project-dir: "\."/);
   assert.match(caller, /@RELEASE_COMMIT_SHA/);
-  assert.match(caller, /checks: read/);
+  assert.match(caller, /^permissions: \{\}$/m);
+  assert.match(caller, /gated-closeout:\n    permissions:\n      contents: write\n      pull-requests: write\n      checks: read\n      actions: write/);
+  assert.match(caller, /required-checks: ""/);
+  assert.match(workflow, /REQUIRED_CHECKS: \$\{\{ inputs.required-checks \}\}/);
   assert.doesNotMatch(caller, /issues: read/);
 });
 
@@ -398,3 +402,33 @@ test("compact PR summary is privacy gated before the bot branch is published", a
   unpublished(f.git);
   assert.equal(existsSync(f.env.PR_STATE!), false);
 }));
+
+for (const mode of ["undeclared", "missing", "neutral", "skipped", "pending", "success", "metadata"]) {
+  test(`required-check metadata boundary ${mode} uses the real workflow and command`, async () => fixture(async f => {
+    f.env.PROTECTION_DENIED = mode === "metadata" ? "false" : "true";
+    // A declared policy never replaces a complete readable policy.
+    f.env.REQUIRED_CHECKS = mode === "undeclared" ? "" : mode === "missing" || mode === "metadata" ? "CI missing" : "CI";
+    if (["neutral", "skipped"].includes(mode)) f.env.HEAD_CONCLUSION = mode;
+    if (mode === "pending") f.env.HEAD_STATUS = "queued";
+    const before = readFileSync(join(f.pmRoot, "tasks", `${f.id}.toon`), "utf8");
+    const result = await f.run([...setup, ...finish]);
+    const report = JSON.parse(readFileSync(join(f.base, "pm-github-closeout.json"), "utf8")) as { closed: number; plan: { code: string }[] };
+    const passes = mode === "success" || mode === "metadata";
+    assert.equal(report.closed, passes ? 1 : 0);
+    assert.equal((await get(f.id, {}, { pmRoot: f.pmRoot, noExtensions: true })).item.status, passes ? "closed" : "in_progress");
+    if (passes) {
+      assert.equal(result.code, 0, result.out);
+      assert.equal(existsSync(f.env.PR_STATE!), true);
+      assert.notEqual(f.git("ls-remote", "origin", `refs/heads/${branch}`), "");
+      const item = (await get(f.id, {}, { pmRoot: f.pmRoot, noExtensions: true })).item;
+      assert.match(String(item.actual_result), mode === "metadata" ? /required checks satisfied/ : /caller-declared and readable required checks satisfied/);
+    } else {
+      assert.equal(report.plan[0]!.code, mode === "undeclared" ? "required_checks_unverified" : "head_checks_failed");
+      if (mode === "undeclared") assert.equal(result.code, 1, result.out);
+      assert.equal(readFileSync(join(f.pmRoot, "tasks", `${f.id}.toon`), "utf8"), before);
+      assert.equal(existsSync(f.env.PR_STATE!), false);
+      unpublished(f.git);
+      assert.doesNotMatch(readFileSync(f.env.RECEIPTS!, "utf8"), /pr (create|edit)|workflow run/);
+    }
+  }));
+}
diff --git a/test/closeout.test.ts b/test/closeout.test.ts
index 07fe32be..a178d668 100644
--- a/test/closeout.test.ts
+++ b/test/closeout.test.ts
@@ -7,7 +7,7 @@ import { join } from "node:path";
 import test from "node:test";
 import { create, close, update, get, config, acquireLock, PmCliError } from "@unbrained/pm-cli/sdk";
 import extension, { fetchJSON, githubApiBase, parseNextLink } from "../index.ts";
-import { CloseoutUsageError, closeoutFailureReason, compactCloseoutReport, declaresDelivery, extractCloseoutItems, runCloseout, type CloseoutApi } from "../closeout.ts";
+import { CloseoutUsageError, closeoutFailureReason, compactCloseoutReport, checkCloseoutHead, declaresDelivery, extractCloseoutItems, runCloseout, type CloseoutApi } from "../closeout.ts";
 import { createExtensionTestHarness } from "@unbrained/pm-cli/sdk/testing";
 import { decode } from "@toon-format/toon";
 import { boundedCli } from "./helpers/bounded-cli.ts";
@@ -280,7 +280,7 @@ test("legacy required statuses use newest context and cannot bypass a failed run
 }));
 
 for (const protection of [403, 404, "empty"] as const) {
-  test(`fallback ${protection} accepts only complete success/neutral/skipped head runs`, async () => tracker(async (root, ids) => {
+  test(`protection ${protection} refuses gaps and preserves readable-empty fallback`, async () => tracker(async (root, ids) => {
     for (const conclusion of ["success", "neutral", "skipped", "failure", null]) {
       const before = snapshot(root);
       await withMockGithub(fixture(link(ids[0]!), (request, response) => {
@@ -288,19 +288,19 @@ for (const protection of [403, 404, "empty"] as const) {
         if (request.url!.includes("/check-runs")) jsonResponse(response, 200, { total_count: 1, check_runs: [run({ conclusion })] });
       }), async () => {
         const report = await closeout(root);
-        assert.equal(report.checks[0]!.rule, "fallback");
-        assert.equal(report.plan[0]!.action, conclusion === "failure" || conclusion === null ? "skip" : "close");
-        assert.match(report.plan[0]!.reason, /Fallback rule/);
+        assert.equal(report.checks[0]!.rule, protection === "empty" ? "fallback" : "unavailable");
+        assert.equal(report.plan[0]!.action, protection !== "empty" || conclusion === "failure" || conclusion === null ? "skip" : "close");
+        assert.equal(report.plan[0]!.code, protection !== "empty" ? "required_checks_unverified" : conclusion === "failure" || conclusion === null ? "head_checks_failed" : "head_checks_passed");
         assert.equal(snapshot(root), before);
       });
     }
   }));
 }
 
-test("fallback rejects empty runs, pending runs and failing legacy statuses", async () => tracker(async (root, ids) => {
+test("readable-empty fallback rejects empty runs, pending runs and failing legacy statuses", async () => tracker(async (root, ids) => {
   for (const mode of ["empty", "pending", "status"]) {
     await withMockGithub(fixture(link(ids[0]!), (request, response) => {
-      if (request.url!.includes("/protection/")) jsonResponse(response, 404, {});
+      if (request.url!.includes("/protection/")) jsonResponse(response, 200, { contexts: [], checks: [] });
       if (request.url!.includes("/check-runs")) jsonResponse(response, 200, { total_count: mode === "empty" ? 0 : 1,
         check_runs: mode === "empty" ? [] : [run({ status: mode === "pending" ? "queued" : "completed" })] });
       if (request.url!.includes("/statuses") && mode === "status") jsonResponse(response, 200, [{ context: "legacy", state: "pending" }]);
@@ -369,7 +369,7 @@ test("invalid selectors fail before HTTP or writes", async () => tracker(async (
   await withMockGithub(fixture(null), async server => {
     for (const options of [{}, { pr: [] }, { pr: false }, { pr: "" }, { pr: ["42", 43] }, { pr: "0" }, { pr: "1.2" }, { pr: "9007199254740992" },
       { pr: "42", since: "2026-10-01" }, { since: "yesterday" }, { since: "2026-02-30" }, { since: "2026-10-01T12:00:00" },
-      { since: "2026-10-01T99:00:00Z" }, { since: false }, { since: "" }, { pr: "42", skip: [] }, { pr: "42", skip: true }]) {
+      { since: "2026-10-01T99:00:00Z" }, { since: false }, { since: "" }, { pr: "42", skip: [] }, { pr: "42", skip: true }, { pr: "42", "required-checks": "  " }, { pr: "42", "required-checks": false }]) {
       await assert.rejects(() => closeout(root, options), CloseoutUsageError);
     }
     for (const repository of [undefined, "invalid", "bad/repo/name", "./repo", "../repo", "fixture/.", "fixture/.."]) await assert.rejects(() => runCloseout(root, repository, { pr: "42" }, api()), CloseoutUsageError);
@@ -570,8 +570,8 @@ test("unreadable effective rules preserve known requirements without hiding auth
       if (request.url!.includes("/rules/branches/")) jsonResponse(response, status, {});
     }), async () => {
       const report = await closeout(root);
-      assert.equal(report.checks[0]!.rule, status === 401 ? "unavailable" : "fallback");
-      assert.equal(report.plan[0]!.action, status === 401 ? "skip" : "close");
+      assert.equal(report.checks[0]!.rule, "unavailable");
+      assert.equal(report.plan[0]!.action, "skip");
     });
   }
 }));
@@ -628,7 +628,7 @@ test("unauthorized classic protection cannot invoke the permissive fallback", as
   });
 }));
 
-test("fallback preserves readable ruleset requirements when classic protection is unavailable", async () => tracker(async (root, ids) => {
+test("declared policy preserves readable ruleset requirements when classic protection is unavailable", async () => tracker(async (root, ids) => {
   for (const conclusion of ["success", "neutral", "skipped", "failure"]) {
     await withMockGithub(fixture(link(ids[0]!), (request, response) => {
       if (request.url!.includes("/protection/")) jsonResponse(response, 403, {});
@@ -637,22 +637,22 @@ test("fallback preserves readable ruleset requirements when classic protection i
       } }]);
       if (request.url!.includes("/check-runs")) jsonResponse(response, 200, { total_count: 1, check_runs: [run({ conclusion })] });
     }), async () => {
-      const report = await closeout(root);
-      assert.equal(report.checks[0]!.rule, "fallback");
+      const report = await closeout(root, { pr: "42", "required-checks": "CI" });
+      assert.equal(report.checks[0]!.rule, "declared");
       assert.equal(report.plan[0]!.action, conclusion === "success" ? "close" : "skip");
     });
   }
 }));
 
-test("unreadable protection sources retain fallback without hiding invalid authorization", async () => tracker(async (root, ids) => {
+test("unreadable protection sources refuse without hiding invalid authorization", async () => tracker(async (root, ids) => {
   for (const status of [403, 404, 401]) {
     await withMockGithub(fixture(link(ids[0]!), (request, response) => {
       if (request.url!.includes("/protection/")) jsonResponse(response, 403, {});
       if (request.url!.includes("/rules/branches/")) jsonResponse(response, status, {});
     }), async () => {
       const report = await closeout(root);
-      assert.equal(report.checks[0]!.rule, status === 401 ? "unavailable" : "fallback");
-      assert.equal(report.plan[0]!.action, status === 401 ? "skip" : "close");
+      assert.equal(report.checks[0]!.rule, "unavailable");
+      assert.equal(report.plan[0]!.action, "skip");
     });
   }
 }));
@@ -692,7 +692,7 @@ test("fallback closes describe the applied check policy and never claim required
     const id = (await create({ title: `Fallback ${conclusion}`, type: "Task", author: "fixture" }, { pmRoot: root, noExtensions: true })).item.id;
     await update(id, { status: "in_progress" }, { pmRoot: root, noExtensions: true });
     await withMockGithub(fixture(link(id), (request, response) => {
-      if (request.url!.includes("/protection/")) jsonResponse(response, 404, {});
+      if (request.url!.includes("/protection/")) jsonResponse(response, 200, { contexts: [], checks: [] });
       if (request.url!.includes("/check-runs")) jsonResponse(response, 200, { total_count: 1, check_runs: [run({ conclusion })] });
     }), async () => {
       assert.equal((await closeout(root, { pr: "42", apply: true })).closed, 1);
@@ -729,3 +729,37 @@ test("compact report counts routine skips and retains individual refusals with t
     assert.equal(full.plan.length, 3);
   });
 }));
+
+test("declared contexts verify exact names and latest statuses without requiring optional success", async () => tracker(async (root, ids) => {
+  await withMockGithub(fixture(link(ids[0]!), (request, response) => {
+    if (request.url!.includes("/protection/") || request.url!.includes("/rules/branches/")) jsonResponse(response, 403, {});
+    if (request.url!.includes("/check-runs")) jsonResponse(response, 200, { total_count: 2, check_runs: [run(), run({ name: "optional", conclusion: "failure" })] });
+    if (request.url!.includes("/statuses")) jsonResponse(response, 200, [{ context: "legacy", state: "success" }, { context: "legacy", state: "failure" }]);
+  }), async () => {
+    const before = snapshot(root);
+    assert.equal((await checkCloseoutHead(api(), repo, pull(link(ids[0]!)))).code, "required_checks_unverified");
+    for (const names of ["CI Legacy", "CI missing"]) {
+      const report = await closeout(root, { pr: "42", apply: true, "required-checks": names });
+      assert.equal(report.closed, 0);
+      assert.equal(report.plan[0]!.code, "head_checks_failed");
+      assert.equal(snapshot(root), before);
+    }
+    const report = await closeout(root, { pr: "42", apply: true, "required-checks": "  CI\tlegacy\nCI  " });
+    assert.equal(report.closed, 1);
+    assert.equal(report.checks[0]!.rule, "declared");
+    assert.deepEqual(report.checks[0]!.passed, ["CI", "legacy"]);
+  });
+}));
+
+test("declared names cannot bypass a readable app binding", async () => tracker(async (root, ids) => {
+  await withMockGithub(fixture(link(ids[0]!), (request, response) => {
+    if (request.url!.includes("/rules/branches/")) jsonResponse(response, 403, {});
+    if (request.url!.includes("/check-runs")) jsonResponse(response, 200, { total_count: 1, check_runs: [run({ app: { id: 8 } })] });
+  }), async () => {
+    const before = snapshot(root);
+    const report = await closeout(root, { pr: "42", apply: true, "required-checks": "CI" });
+    assert.equal(report.closed, 0);
+    assert.equal(report.plan[0]!.code, "head_checks_failed");
+    assert.equal(snapshot(root), before);
+  });
+}));

From 6e4d86e1e3e61092e8927d094d76d91a295d82da Mon Sep 17 00:00:00 2001
From: SteveBot <1153461+unbraind@users.noreply.github.com>
Date: Fri, 16 Oct 2026 21:53:52 +0200
Subject: [PATCH 2/3] fix(closeout): scope declared-check success evidence

---
 .agents/pm/history/pm-github-f1ev.jsonl | 6 ++++++
 .agents/pm/issues/pm-github-f1ev.toon   | 8 ++++++--
 README.md                               | 9 ++++++---
 closeout.ts                             | 9 +++++----
 test/closeout-workflow.test.ts          | 2 +-
 test/closeout.test.ts                   | 2 ++
 6 files changed, 26 insertions(+), 10 deletions(-)

diff --git a/.agents/pm/history/pm-github-f1ev.jsonl b/.agents/pm/history/pm-github-f1ev.jsonl
index 520b4745..16e246eb 100644
--- a/.agents/pm/history/pm-github-f1ev.jsonl
+++ b/.agents/pm/history/pm-github-f1ev.jsonl
@@ -20,3 +20,9 @@
 {"hash_algorithm":"sha256","ts":"2026-10-16T17:41:09.873Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/9","value":{"created_at":"2026-10-16T17:41:09.873Z","author":"codex-agent","text":"Rebased acceptance on main 6d9cc23: npm run release:check passed 747/747 tests with zero failures/skips/cancellations/todos. Exact coverage remains 13 modules, 10082/10082 lines and statements, 256/256 functions, 2858/2858 branches (100/100/100/100; thresholds unchanged). All release gates passed. Dedicated integration checks also passed: node --test test/self-versioned-workflows.test.ts (36/36) and node --test --test-name-pattern='closeout contract' test/closeout-workflow.test.ts (1/1). The rebased Bun gate is now running."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T17:41:09.873Z"}],"before_hash":"f6d98341e2f3e8b7794a72bd2b2a4d964ad957884ffd8cb1455a71649bd8653a","after_hash":"3d91279c7db59105c0a18c8cca829ea0d85e7e78626f392452d01f00515c4a83","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"9b10d8ca06009ddbe49ee6cb9b12827f87b98dabb45e35d9987cd3605bd75527"}
 {"hash_algorithm":"sha256","ts":"2026-10-16T18:06:25.089Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/10","value":{"created_at":"2026-10-16T18:06:25.089Z","author":"codex-agent","text":"Final rebased acceptance: npm run release:check and bun run release:check both passed 747/747 tests, zero failures/skips/cancellations/todos, exact 100/100/100/100 coverage across 13 authored modules (10082 lines/statements, 256 functions, 2858 branches). All release gates passed at unchanged thresholds on the implementation rebased over #139. Behavioral head-function and workflow-guard revert proofs remain recorded above; real-command boundary tests passed 9/9. Ready for review; no release, merge, downstream adoption or item close is performed. Caller-declared context lists must stay aligned with required checks."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T18:06:25.089Z"}],"before_hash":"3d91279c7db59105c0a18c8cca829ea0d85e7e78626f392452d01f00515c4a83","after_hash":"a30a356d83e28ea153b2339786a91336952cfcbade9446f834bee5e98f584fff","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"e81721bfbfc62b985ed775994060faa16260233d84902a1b53e1d9ac11b07325"}
 {"hash_algorithm":"sha256","ts":"2026-10-16T18:06:26.222Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e532c36ed40fc318cac03fae","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"release","patch":[{"op":"remove","path":"/metadata/claim_principal"},{"op":"remove","path":"/metadata/assignee"},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T18:06:26.222Z"}],"before_hash":"a30a356d83e28ea153b2339786a91336952cfcbade9446f834bee5e98f584fff","after_hash":"95a01ca6cd3e4bc3e6fcaec8f066166e9dd8fc809ad76ff49c536c9b22248393","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"2277a5f3985a79d596dcc9aa2cae1e8d4291368307da1269f08ad01777e82515"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T19:13:41.651Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"claim","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:13:41.651Z"},{"op":"add","path":"/metadata/assignee","value":"codex-agent"},{"op":"add","path":"/metadata/claim_principal","value":"codex-agent"}],"before_hash":"95a01ca6cd3e4bc3e6fcaec8f066166e9dd8fc809ad76ff49c536c9b22248393","after_hash":"cfb10323617ac6fc461258cf1505a0947b7b568736d0f398a7917e2a8607d2f0","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"8569d136e977d00cf8c667e6f507a42adc3a735ac8a7c7bc27f3b4bfac1b1191"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T19:13:43.207Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/11","value":{"created_at":"2026-10-16T19:13:43.207Z","author":"codex-agent","text":"Round 1: address declared-policy success wording. Keep head_checks_passed and rule=declared stable; describe only caller-declared success and unavailable repository required-check metadata in the gate reason and structured evidence. Update README and real-workflow assertions, prove failures by reverting each changed function separately, then run both release gates at unchanged thresholds. The branch already contains origin/main; no merge is needed."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:13:43.207Z"}],"before_hash":"cfb10323617ac6fc461258cf1505a0947b7b568736d0f398a7917e2a8607d2f0","after_hash":"1430676a0c558842bbe2fce0d9613199b7319d7af39a5003b94b8324c50bf680","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"4b600e1bdbcb8266b2ae14454b59b3fc4d0604aaf0884efa54def7d5fdab004e"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T19:15:42.418Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/12","value":{"created_at":"2026-10-16T19:15:42.418Z","author":"codex-agent","text":"Round 1 behavioral revert proof: restoring only checkCloseoutHead from the pre-review HEAD, rebuilding, and running node --test --test-name-pattern='declared contexts' test/closeout.test.ts fails exactly 1/1 test on the old overstated gate reason. Restoring only checkEvidence, rebuilding, and running node --test --test-name-pattern='required-check metadata boundary success' test/closeout-workflow.test.ts fails exactly 1/1 real-workflow test on persisted actual_result. Each function was restored after its proof and the fixed build succeeds. Pinned pm test pm-github-f1ev --run --progress passed all 9/9 linked boundary tests. Codes remain head_checks_passed with rule=declared; README documents this contract."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:15:42.418Z"}],"before_hash":"1430676a0c558842bbe2fce0d9613199b7319d7af39a5003b94b8324c50bf680","after_hash":"6832680e61b1720338454eddea260026777e54ab205482e92de96362157f8d6f","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"26ae5c73aa02c6edbeba09843339d7f8e2247d7f240c84e78213b0e3875b6f73"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T19:35:17.666Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/13","value":{"created_at":"2026-10-16T19:35:17.666Z","author":"codex-agent","text":"Round 1 npm acceptance: CI-identity-isolated npm run release:check passed 747/747 tests with zero failures/skips/cancellations/todos. Coverage is exact 100/100/100/100 across 13 authored modules: 10083/10083 lines and statements, 256/256 functions, 2858/2858 branches. All release gates passed at unchanged thresholds. CI-style pinned tracker health also passes; normal-home health reports local extension host version skew, with no tracked integrity/history drift. Bun release gate is running next."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:35:17.666Z"}],"before_hash":"6832680e61b1720338454eddea260026777e54ab205482e92de96362157f8d6f","after_hash":"1b98d9704f9268e4ca6a992b3bc0fd1b71e50cd0bd1a7704bc9fe3a77deca05d","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"7eb74ca41d0abb069c99d59f7ee0bbc663cddb30a026c101f32cc06dbe5ff0d0"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T19:53:20.313Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/14","value":{"created_at":"2026-10-16T19:53:20.313Z","author":"codex-agent","text":"Round 1 final acceptance: npm run release:check and bun run release:check both passed 747/747 tests with zero failures/skips/cancellations/todos using isolated CI-style Git identity. Both measured exact 100/100/100/100 coverage across 13 authored modules (10083 lines/statements, 256 functions, 2858 branches) at unchanged thresholds, with all static, privacy, audit, package, changelog and attestation gates passing. The fixed linked boundary suite passed 9/9; independent function revert proofs are recorded above. Declared-only success now explicitly states that repository required-check metadata was unavailable, without changing head_checks_passed or rule=declared. Caller declarations still require maintenance to match repository policy. Ready to push PR #140; item remains in progress for orchestrator verification."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:53:20.313Z"}],"before_hash":"1b98d9704f9268e4ca6a992b3bc0fd1b71e50cd0bd1a7704bc9fe3a77deca05d","after_hash":"8bf812069d3ad2a77a214a92ae6d049f2ca5e26bacbbf9f7c4df895d69b54e2f","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"c9d989fdfc80fb2758d77ad9643a87648049b0c903892002456e080e037a3480"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T19:53:20.961Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"release","patch":[{"op":"remove","path":"/metadata/claim_principal"},{"op":"remove","path":"/metadata/assignee"},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:53:20.961Z"}],"before_hash":"8bf812069d3ad2a77a214a92ae6d049f2ca5e26bacbbf9f7c4df895d69b54e2f","after_hash":"b31d52ee42c42d4d289e0acc43fc42a67a17b25d1d528843020908a094a41cf6","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"19987d0b0e02a77f99bf89d8b59fc45514725b515808a4c44c6dd4b63857ee13"}
diff --git a/.agents/pm/issues/pm-github-f1ev.toon b/.agents/pm/issues/pm-github-f1ev.toon
index 15bffe03..4e2176f8 100644
--- a/.agents/pm/issues/pm-github-f1ev.toon
+++ b/.agents/pm/issues/pm-github-f1ev.toon
@@ -6,10 +6,10 @@ status: in_progress
 priority: 1
 tags[2]: closeout,validation
 created_at: "2026-10-16T16:04:00.890Z"
-updated_at: "2026-10-16T18:06:26.222Z"
+updated_at: "2026-10-16T19:53:20.961Z"
 author: codex-agent
 acceptance_criteria: Unavailable metadata without declarations never closes or publishes; declared contexts require success; metadata-backed behavior remains unchanged; npm and Bun release gates pass.
-comments[11]{created_at,author,text}:
+comments[15]{created_at,author,text}:
   "2026-10-16T16:04:15.081Z",codex-agent,"Plan: distinguish unavailable protection metadata from a readable empty policy. Refuse with required_checks_unverified unless required-checks declares contexts; retain readable app-bound requirements. Exercise real command and workflow shell against injected HTTP evidence and real trackers. Keep workflow edits minimal around open PR #139."
   "2026-10-16T16:08:21.292Z",codex-agent,"Implementation distinguishes metadata gaps from a readable empty policy. required_checks_unverified blocks workflow publication; required-checks verifies declared contexts plus readable app-bound requirements. Regression coverage uses actual workflow bash, a packed extension, injected HTTP APIs, real SDK tracker mutations and an isolated Git remote. Caller write scopes now belong only to gated-closeout."
   "2026-10-16T16:10:06.013Z",codex-agent,"Behavioural revert proof: restored only checkCloseoutHead to its previous implementation (retaining the added argument for compilation), rebuilt dist, and ran node --test --test-name-pattern='required-check metadata boundary (undeclared|missing) ' test/closeout-workflow.test.ts. Both tests failed: each observed closed=1 where closed=0 was required (0 passed, 2 failed). Fixed function restored and rebuilt. This exercises the packed command and real workflow rather than SDK mocks."
@@ -21,6 +21,10 @@ comments[11]{created_at,author,text}:
   "2026-10-16T17:05:41.112Z",codex-agent,PR #139 merged during validation. Rebased onto main commit 6d9cc23 and resolved the caller-template conflict by preserving the self-versioned release SHA/tag comment plus this change's job-only permissions and required-checks input. No self-versioning implementation was changed. Previous 711/711 npm and Bun receipts predate this integration; rerunning the full gates on the rebased branch before delivery.
   "2026-10-16T17:41:09.873Z",codex-agent,"Rebased acceptance on main 6d9cc23: npm run release:check passed 747/747 tests with zero failures/skips/cancellations/todos. Exact coverage remains 13 modules, 10082/10082 lines and statements, 256/256 functions, 2858/2858 branches (100/100/100/100; thresholds unchanged). All release gates passed. Dedicated integration checks also passed: node --test test/self-versioned-workflows.test.ts (36/36) and node --test --test-name-pattern='closeout contract' test/closeout-workflow.test.ts (1/1). The rebased Bun gate is now running."
   "2026-10-16T18:06:25.089Z",codex-agent,"Final rebased acceptance: npm run release:check and bun run release:check both passed 747/747 tests, zero failures/skips/cancellations/todos, exact 100/100/100/100 coverage across 13 authored modules (10082 lines/statements, 256 functions, 2858 branches). All release gates passed at unchanged thresholds on the implementation rebased over #139. Behavioral head-function and workflow-guard revert proofs remain recorded above; real-command boundary tests passed 9/9. Ready for review; no release, merge, downstream adoption or item close is performed. Caller-declared context lists must stay aligned with required checks."
+  "2026-10-16T19:13:43.207Z",codex-agent,"Round 1: address declared-policy success wording. Keep head_checks_passed and rule=declared stable; describe only caller-declared success and unavailable repository required-check metadata in the gate reason and structured evidence. Update README and real-workflow assertions, prove failures by reverting each changed function separately, then run both release gates at unchanged thresholds. The branch already contains origin/main; no merge is needed."
+  "2026-10-16T19:15:42.418Z",codex-agent,"Round 1 behavioral revert proof: restoring only checkCloseoutHead from the pre-review HEAD, rebuilding, and running node --test --test-name-pattern='declared contexts' test/closeout.test.ts fails exactly 1/1 test on the old overstated gate reason. Restoring only checkEvidence, rebuilding, and running node --test --test-name-pattern='required-check metadata boundary success' test/closeout-workflow.test.ts fails exactly 1/1 real-workflow test on persisted actual_result. Each function was restored after its proof and the fixed build succeeds. Pinned pm test pm-github-f1ev --run --progress passed all 9/9 linked boundary tests. Codes remain head_checks_passed with rule=declared; README documents this contract."
+  "2026-10-16T19:35:17.666Z",codex-agent,"Round 1 npm acceptance: CI-identity-isolated npm run release:check passed 747/747 tests with zero failures/skips/cancellations/todos. Coverage is exact 100/100/100/100 across 13 authored modules: 10083/10083 lines and statements, 256/256 functions, 2858/2858 branches. All release gates passed at unchanged thresholds. CI-style pinned tracker health also passes; normal-home health reports local extension host version skew, with no tracked integrity/history drift. Bun release gate is running next."
+  "2026-10-16T19:53:20.313Z",codex-agent,"Round 1 final acceptance: npm run release:check and bun run release:check both passed 747/747 tests with zero failures/skips/cancellations/todos using isolated CI-style Git identity. Both measured exact 100/100/100/100 coverage across 13 authored modules (10083 lines/statements, 256 functions, 2858 branches) at unchanged thresholds, with all static, privacy, audit, package, changelog and attestation gates passing. The fixed linked boundary suite passed 9/9; independent function revert proofs are recorded above. Declared-only success now explicitly states that repository required-check metadata was unavailable, without changing head_checks_passed or rule=declared. Caller declarations still require maintenance to match repository policy. Ready to push PR #140; item remains in progress for orchestrator verification."
 files[5]{path,scope}:
   .github/workflows/pm-github-closeout.yml,project
   closeout.ts,project
diff --git a/README.md b/README.md
index 57f9fbb2..5a6f9d66 100644
--- a/README.md
+++ b/README.md
@@ -90,7 +90,9 @@ plan row and selected PR check. Global `--json` always returns the full shape
 `{item, pr, action: "close" | "skip", code, reason}`. Codes distinguish
 `head_evidence_unavailable`, `incomplete_merge`, `head_checks_failed`, and
 `head_checks_passed` without parsing explanations. `checks` reports each PR's rule,
-verdict, explanation and successful check names. Dry-run rows describe intended
+verdict, explanation and successful check names. `head_checks_passed` remains the success
+code for all passing policies; `rule: "declared"` proves the caller-declared contexts,
+while repository required-check metadata was unavailable. Dry-run rows describe intended
 closes; `closed` counts writes only with `--apply`. `--dry-run` overrides
 `--apply`. All API planning finishes before the first SDK close. Individual SDK
 refusals become skip rows containing a diagnostic code (or a generic failure
@@ -113,8 +115,9 @@ structured close fields it supports, without overriding tracker policy:
 - `actual_result`: `Merged at <time>; required checks satisfied; successful head checks:
   <names>` when branch protection or rulesets name required checks (the list holds every
   successful head check, optional ones included). Under the declared policy it reads
-  `Merged at <time>; caller-declared and readable required checks satisfied; successful
-  head checks: <names>`. Under the fallback rule (fully readable, empty policy) it reads
+  `Merged at <time>; caller-declared checks concluded success; repository required-check
+  metadata was unavailable; successful head checks: <names>`. Under the fallback rule
+  (fully readable, empty policy) it reads
   `Merged at <time>; no readable required checks, so every
   completed head check run had to succeed, be neutral or be skipped; successful checks:
   <names>`, or `…; no check reported success` when only neutral or skipped runs exist.
diff --git a/closeout.ts b/closeout.ts
index 2682683d..53b178bf 100644
--- a/closeout.ts
+++ b/closeout.ts
@@ -223,7 +223,7 @@ export async function checkCloseoutHead(api: CloseoutApi, repo: string, pr: Pull
       }
     }
     if (!policy.fallback) {
-      result.reason = policy.unavailable ? "Caller-declared and readable required head checks concluded success."
+      result.reason = policy.unavailable ? "Caller-declared checks concluded success; repository required-check metadata was unavailable."
         : "Required head checks concluded success (branch-protection and ruleset rule).";
     } else {
       if (runs.length === 0) { result.reason = "Fallback rule: no check runs at head; evidence missing."; return result; }
@@ -336,13 +336,14 @@ function usage<T>(parse: () => T): T {
 /**
  * Describe the verified check evidence for a passed gate without overstating it.
  *
- * The required and declared rules state that their required checks were satisfied and list every successful head check (optional
- * ones included). The fallback rule (fully readable metadata with no required contexts)
+ * The required rule reports required-check success; the declared rule reports caller-declared success
+ * and unavailable repository metadata. Both list every successful head check (optional ones included).
+ * The fallback rule (fully readable metadata with no required contexts)
  * may pass on neutral or skipped runs alone, so it states that policy and lists successful checks only when any exist.
  */
 export function checkEvidence(gate: CheckGate): string {
   // gate.passed holds every successful head check, optional ones included, so it is never labelled as the required set.
-  if (gate.rule === "declared") return `caller-declared and readable required checks satisfied; successful head checks: ${gate.passed.join(", ")}`;
+  if (gate.rule === "declared") return `caller-declared checks concluded success; repository required-check metadata was unavailable; successful head checks: ${gate.passed.join(", ")}`;
   if (gate.rule === "required") return `required checks satisfied; successful head checks: ${gate.passed.join(", ")}`;
   const successes = gate.passed.length > 0 ? `successful checks: ${gate.passed.join(", ")}` : "no check reported success";
   return `no readable required checks, so every completed head check run had to succeed, be neutral or be skipped; ${successes}`;
diff --git a/test/closeout-workflow.test.ts b/test/closeout-workflow.test.ts
index 0b8554f3..bd8024e0 100644
--- a/test/closeout-workflow.test.ts
+++ b/test/closeout-workflow.test.ts
@@ -421,7 +421,7 @@ for (const mode of ["undeclared", "missing", "neutral", "skipped", "pending", "s
       assert.equal(existsSync(f.env.PR_STATE!), true);
       assert.notEqual(f.git("ls-remote", "origin", `refs/heads/${branch}`), "");
       const item = (await get(f.id, {}, { pmRoot: f.pmRoot, noExtensions: true })).item;
-      assert.match(String(item.actual_result), mode === "metadata" ? /required checks satisfied/ : /caller-declared and readable required checks satisfied/);
+      assert.match(String(item.actual_result), mode === "metadata" ? /required checks satisfied/ : /caller-declared checks concluded success; repository required-check metadata was unavailable; successful head checks: CI/);
     } else {
       assert.equal(report.plan[0]!.code, mode === "undeclared" ? "required_checks_unverified" : "head_checks_failed");
       if (mode === "undeclared") assert.equal(result.code, 1, result.out);
diff --git a/test/closeout.test.ts b/test/closeout.test.ts
index a178d668..65771419 100644
--- a/test/closeout.test.ts
+++ b/test/closeout.test.ts
@@ -747,6 +747,8 @@ test("declared contexts verify exact names and latest statuses without requiring
     const report = await closeout(root, { pr: "42", apply: true, "required-checks": "  CI\tlegacy\nCI  " });
     assert.equal(report.closed, 1);
     assert.equal(report.checks[0]!.rule, "declared");
+    assert.equal(report.checks[0]!.code, "head_checks_passed");
+    assert.equal(report.checks[0]!.reason, "Caller-declared checks concluded success; repository required-check metadata was unavailable.");
     assert.deepEqual(report.checks[0]!.passed, ["CI", "legacy"]);
   });
 }));

From 35f907f0795aa5ff7d1c55aedf4ea8f1dfad7a2e Mon Sep 17 00:00:00 2001
From: SteveBot <1153461+unbraind@users.noreply.github.com>
Date: Fri, 16 Oct 2026 23:41:59 +0200
Subject: [PATCH 3/3] fix(closeout): preserve newline-delimited required
 context names

---
 .agents/pm/history/pm-github-f1ev.jsonl  | 10 +++++++
 .agents/pm/issues/pm-github-f1ev.toon    | 31 +++++++++++++++++---
 .github/workflows/pm-github-closeout.yml |  3 +-
 README.md                                | 21 ++++++++++----
 closeout.ts                              | 11 +++++--
 docs/closeout-workflow-caller.yml        |  5 +++-
 index.ts                                 |  2 +-
 test/closeout-workflow.test.ts           | 37 ++++++++++++++++++++----
 test/closeout.test.ts                    | 26 +++++++++++++++--
 9 files changed, 123 insertions(+), 23 deletions(-)

diff --git a/.agents/pm/history/pm-github-f1ev.jsonl b/.agents/pm/history/pm-github-f1ev.jsonl
index 16e246eb..5ae36432 100644
--- a/.agents/pm/history/pm-github-f1ev.jsonl
+++ b/.agents/pm/history/pm-github-f1ev.jsonl
@@ -26,3 +26,13 @@
 {"hash_algorithm":"sha256","ts":"2026-10-16T19:35:17.666Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/13","value":{"created_at":"2026-10-16T19:35:17.666Z","author":"codex-agent","text":"Round 1 npm acceptance: CI-identity-isolated npm run release:check passed 747/747 tests with zero failures/skips/cancellations/todos. Coverage is exact 100/100/100/100 across 13 authored modules: 10083/10083 lines and statements, 256/256 functions, 2858/2858 branches. All release gates passed at unchanged thresholds. CI-style pinned tracker health also passes; normal-home health reports local extension host version skew, with no tracked integrity/history drift. Bun release gate is running next."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:35:17.666Z"}],"before_hash":"6832680e61b1720338454eddea260026777e54ab205482e92de96362157f8d6f","after_hash":"1b98d9704f9268e4ca6a992b3bc0fd1b71e50cd0bd1a7704bc9fe3a77deca05d","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"7eb74ca41d0abb069c99d59f7ee0bbc663cddb30a026c101f32cc06dbe5ff0d0"}
 {"hash_algorithm":"sha256","ts":"2026-10-16T19:53:20.313Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/14","value":{"created_at":"2026-10-16T19:53:20.313Z","author":"codex-agent","text":"Round 1 final acceptance: npm run release:check and bun run release:check both passed 747/747 tests with zero failures/skips/cancellations/todos using isolated CI-style Git identity. Both measured exact 100/100/100/100 coverage across 13 authored modules (10083 lines/statements, 256 functions, 2858 branches) at unchanged thresholds, with all static, privacy, audit, package, changelog and attestation gates passing. The fixed linked boundary suite passed 9/9; independent function revert proofs are recorded above. Declared-only success now explicitly states that repository required-check metadata was unavailable, without changing head_checks_passed or rule=declared. Caller declarations still require maintenance to match repository policy. Ready to push PR #140; item remains in progress for orchestrator verification."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:53:20.313Z"}],"before_hash":"1b98d9704f9268e4ca6a992b3bc0fd1b71e50cd0bd1a7704bc9fe3a77deca05d","after_hash":"8bf812069d3ad2a77a214a92ae6d049f2ca5e26bacbbf9f7c4df895d69b54e2f","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"c9d989fdfc80fb2758d77ad9643a87648049b0c903892002456e080e037a3480"}
 {"hash_algorithm":"sha256","ts":"2026-10-16T19:53:20.961Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"e7be906c617b96f656c65c8d","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"release","patch":[{"op":"remove","path":"/metadata/claim_principal"},{"op":"remove","path":"/metadata/assignee"},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T19:53:20.961Z"}],"before_hash":"8bf812069d3ad2a77a214a92ae6d049f2ca5e26bacbbf9f7c4df895d69b54e2f","after_hash":"b31d52ee42c42d4d289e0acc43fc42a67a17b25d1d528843020908a094a41cf6","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"19987d0b0e02a77f99bf89d8b59fc45514725b515808a4c44c6dd4b63857ee13"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T20:39:24.167Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"claim","patch":[{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T20:39:24.167Z"},{"op":"add","path":"/metadata/assignee","value":"codex-agent"},{"op":"add","path":"/metadata/claim_principal","value":"codex-agent"}],"before_hash":"b31d52ee42c42d4d289e0acc43fc42a67a17b25d1d528843020908a094a41cf6","after_hash":"96f9ccdc59f90c704d1f476a805fbb324745860b839ddad634f4f24c854dfec5","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"04624b0057c100edf5751f7091b49bf89d013df5e519b8c367ad634059e72371"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T20:39:26.384Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/15","value":{"created_at":"2026-10-16T20:39:26.384Z","author":"codex-agent","text":"Round 2: replace whitespace splitting with one trimmed required context per line, dropping empty lines. Preserve the quoted newline-joined workflow argument, update CLI help and caller examples, and exercise exact names plus literal shell-shaped names through real workflow bash and SDK trackers. Record behavioural revert proof, then run unchanged npm and Bun release gates."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T20:39:26.384Z"}],"before_hash":"96f9ccdc59f90c704d1f476a805fbb324745860b839ddad634f4f24c854dfec5","after_hash":"1a6482b6e965e510bb0b4a4930e2ade1546e817461489000e7b19f66524ddc71","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"bb051df2d03aab6b46ac282634772385215cabcce3006d3b45f7f80133e9ef48"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T20:40:51.590Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"tests_add","patch":[{"op":"add","path":"/metadata/tests/1","value":{"command":"npm run build && npm run build:test && node --test --test-name-pattern=\"newline|required-check metadata boundary|declared contexts|declared names|invalid selectors|closeout contract\" test/closeout.test.ts test/closeout-workflow.test.ts","scope":"project","provenance":{"author":"codex-agent","created_at":"2026-10-16T20:40:51.494Z","source_kind":"local_mutation","source_ref":"fix/closeout-required-checks-unverified"}}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T20:40:51.590Z"}],"before_hash":"1a6482b6e965e510bb0b4a4930e2ade1546e817461489000e7b19f66524ddc71","after_hash":"f1d1f716a35f19fd77a775cfab637d5732da9e3a55ed7f4b9ac4f213342c7921","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"27ec28cbf6de16e38f3fe44a2f518d1e5a4da193582799a786d5a84e63f7e784"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T20:43:48.686Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/16","value":{"created_at":"2026-10-16T20:43:48.686Z","author":"codex-agent","text":"Round 2 behavioural revert proof: restored only the required-context parsing stanza inside runCloseout to the pre-review repeated(...).flatMap(value => value.split(/\\s+/)) implementation, retaining all new tests and other changes. After npm run build, node --test --test-name-pattern=\"newline required contexts\" test/closeout-workflow.test.ts failed all 5/5 tests (0 passed; closed=0 instead of 1): single spaced name, spaces/parentheses/slashes, fleet example, empty lines/trimming, and literal shell-shaped names. Restored the line parser and rebuilt successfully. The fixed focused suite passed 18/18 with no skips or cancellations."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T20:43:48.686Z"}],"before_hash":"f1d1f716a35f19fd77a775cfab637d5732da9e3a55ed7f4b9ac4f213342c7921","after_hash":"40c4fbc677461c2102fa141a3a6d25b3fb1f7cb1a8ad125c31ab829f45739dd9","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"67416179887e9e921b5b90479d2d74120cd7dc1330efca84d863cff2c71c7680"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T20:44:12.196Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"tests_update","patch":[{"op":"add","path":"/metadata/tests/1/timeout_seconds","value":1800},{"op":"add","path":"/metadata/tests/1/updated_by","value":"codex-agent"},{"op":"add","path":"/metadata/tests/1/updated_at","value":"2026-10-16T20:44:12.196Z"},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T20:44:12.196Z"}],"before_hash":"40c4fbc677461c2102fa141a3a6d25b3fb1f7cb1a8ad125c31ab829f45739dd9","after_hash":"2c382f4cdb54e259b5f6ab9d930e35e0b53df5c8b0f47eed8b02bac72aed6ffa","item_hash_version":3,"context":{"linked_test_update":{"index":2,"fields":["timeout_seconds"]},"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"3572df872367b9968db6981219887bc8758046df07d19cdb4831f79e8cf410ee"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T20:46:05.571Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/17","value":{"created_at":"2026-10-16T20:46:05.570Z","author":"codex-agent","text":"The first pinned linked-test run hit its already-loaded default 120-second timeout before completion, with no assertion failures in completed cases. Set the new linked entry to timeout_seconds=1800 and rerun with an explicit --timeout 1800. This timeout is missing acceptance evidence, not a pass."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T20:46:05.571Z"}],"before_hash":"2c382f4cdb54e259b5f6ab9d930e35e0b53df5c8b0f47eed8b02bac72aed6ffa","after_hash":"40e7a3492a3e349c04983ddbd879401bc7b78b025b86b89a0ee3c257c7978cb1","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"c7d9c5c1119664e2f7cd0261163f5af09915b2f68d3f1129b3ec7098bf76fa78"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T20:49:00.145Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/18","value":{"created_at":"2026-10-16T20:49:00.145Z","author":"codex-agent","text":"Pinned node_modules/.bin/pm test pm-github-f1ev --run --progress --only-last --timeout 1800 passed the new linked command: 18/18 tests, zero failures/skips/cancellations/todos, no timeout. Test-result tracking is disabled by project policy, so this comment preserves the completed acceptance receipt without changing that policy."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T20:49:00.145Z"}],"before_hash":"40e7a3492a3e349c04983ddbd879401bc7b78b025b86b89a0ee3c257c7978cb1","after_hash":"16632cdf942098cae975b179f9d6206918adbac454133b3f3a70731e7fd1eea2","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"ec8ebe470259e2f2a20459084c62a1758b6ea6fb2cc1fb9fdd149912f017a0da"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T21:13:33.378Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/19","value":{"created_at":"2026-10-16T21:13:33.378Z","author":"codex-agent","text":"Round 2 npm run release:check passed with isolated CI-style Git identity: 754/754 tests, zero failures/skips/cancellations/todos. Exact 100/100/100/100 coverage across 13 authored modules: 10090/10090 lines and statements, 257/257 functions, 2866/2866 branches. All static, privacy, audit, package, changelog and attestation gates passed at unchanged thresholds. Tracker health passed with advisory telemetry/stale-item findings. Starting the Bun release gate on unchanged code."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T21:13:33.378Z"}],"before_hash":"16632cdf942098cae975b179f9d6206918adbac454133b3f3a70731e7fd1eea2","after_hash":"602c068bd3312a36ca85bbe21c797864fbe359c36320896e9cf28e10180e5002","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"ff52c0afc028967022f3f6defc49ab85eb841067d8252ff8be4b0cb0e12dd916"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T21:41:47.426Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":null,"topic":null},"op":"comment_add","patch":[{"op":"add","path":"/metadata/comments/20","value":{"created_at":"2026-10-16T21:41:47.426Z","author":"codex-agent","text":"Round 2 final local acceptance: npm run release:check and bun run release:check both passed 754/754 tests with zero failures/skips/cancellations/todos using isolated CI-style Git identity. Both measured exact 100/100/100/100 coverage across 13 authored modules (10090 lines/statements, 257 functions, 2866 branches), at unchanged thresholds; all static, privacy, audit, package, changelog and attestation gates passed. The pinned linked suite passed 18/18 after replacing its default timeout with 1800 seconds. The 5/5 failing behavioural revert proof is recorded above. required-checks now preserves one trimmed literal context per line, drops blank lines, and retains safe quoted transport. README, CLI help and caller template show the new contract, including test (22) and test (26). No new PM CLI/SDK defect was identified. Caller declarations must still match repository policy. Ready to push PR #140; GitHub checks will be polled without thread replies or mentions. Item remains in progress for orchestrator verification."}},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T21:41:47.426Z"}],"before_hash":"602c068bd3312a36ca85bbe21c797864fbe359c36320896e9cf28e10180e5002","after_hash":"c850b6473c5e07abce82499251551313871768a892e42c09fc6eba06c9d74a4d","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"substantive","record_hash_version":1,"record_hash":"d06edd145315f7cc8ba40a99baa118a6f870a8adf6cccdc2102cf0df59d94055"}
+{"hash_algorithm":"sha256","ts":"2026-10-16T21:41:48.600Z","author":"codex-agent","author_source":"asserted","agent_harness":"codex","agent_model":"gpt-6.1-sol","agent_model_source":"probe","agent_instance":"a389994a963e77cdce5c0ddc","agent_provenance":{"model":{"value":"gpt-6.1-sol","source":"probe"},"effort":{"value":"high","source":"probe"},"role":{"value":"implementer","source":"argv"},"topic":null},"op":"release","patch":[{"op":"remove","path":"/metadata/claim_principal"},{"op":"remove","path":"/metadata/assignee"},{"op":"replace","path":"/metadata/updated_at","value":"2026-10-16T21:41:48.600Z"}],"before_hash":"c850b6473c5e07abce82499251551313871768a892e42c09fc6eba06c9d74a4d","after_hash":"688cad420842cd7c346f96f3ee3c49797fe0844c88772ffa1461a9cbb82099a0","item_hash_version":3,"context":{"agent_provenance_outcomes":{"model":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"effort":{"status":"resolved","resolver":"codex_session_file","rule_version":"v1"},"version":{"status":"unavailable","reason":"harness_unavailable","resolver":"ai_agent_version","rule_version":"v1"}}},"event_class":"maintenance","record_hash_version":1,"record_hash":"ad51e72d61502c51ed7f29a059a915421bc570f8e6279aa03b4ef846631563b3"}
diff --git a/.agents/pm/issues/pm-github-f1ev.toon b/.agents/pm/issues/pm-github-f1ev.toon
index 4e2176f8..d63758f5 100644
--- a/.agents/pm/issues/pm-github-f1ev.toon
+++ b/.agents/pm/issues/pm-github-f1ev.toon
@@ -6,10 +6,10 @@ status: in_progress
 priority: 1
 tags[2]: closeout,validation
 created_at: "2026-10-16T16:04:00.890Z"
-updated_at: "2026-10-16T19:53:20.961Z"
+updated_at: "2026-10-16T21:41:48.600Z"
 author: codex-agent
 acceptance_criteria: Unavailable metadata without declarations never closes or publishes; declared contexts require success; metadata-backed behavior remains unchanged; npm and Bun release gates pass.
-comments[15]{created_at,author,text}:
+comments[21]{created_at,author,text}:
   "2026-10-16T16:04:15.081Z",codex-agent,"Plan: distinguish unavailable protection metadata from a readable empty policy. Refuse with required_checks_unverified unless required-checks declares contexts; retain readable app-bound requirements. Exercise real command and workflow shell against injected HTTP evidence and real trackers. Keep workflow edits minimal around open PR #139."
   "2026-10-16T16:08:21.292Z",codex-agent,"Implementation distinguishes metadata gaps from a readable empty policy. required_checks_unverified blocks workflow publication; required-checks verifies declared contexts plus readable app-bound requirements. Regression coverage uses actual workflow bash, a packed extension, injected HTTP APIs, real SDK tracker mutations and an isolated Git remote. Caller write scopes now belong only to gated-closeout."
   "2026-10-16T16:10:06.013Z",codex-agent,"Behavioural revert proof: restored only checkCloseoutHead to its previous implementation (retaining the added argument for compilation), rebuilt dist, and ran node --test --test-name-pattern='required-check metadata boundary (undeclared|missing) ' test/closeout-workflow.test.ts. Both tests failed: each observed closed=1 where closed=0 was required (0 passed, 2 failed). Fixed function restored and rebuilt. This exercises the packed command and real workflow rather than SDK mocks."
@@ -25,14 +25,37 @@ comments[15]{created_at,author,text}:
   "2026-10-16T19:15:42.418Z",codex-agent,"Round 1 behavioral revert proof: restoring only checkCloseoutHead from the pre-review HEAD, rebuilding, and running node --test --test-name-pattern='declared contexts' test/closeout.test.ts fails exactly 1/1 test on the old overstated gate reason. Restoring only checkEvidence, rebuilding, and running node --test --test-name-pattern='required-check metadata boundary success' test/closeout-workflow.test.ts fails exactly 1/1 real-workflow test on persisted actual_result. Each function was restored after its proof and the fixed build succeeds. Pinned pm test pm-github-f1ev --run --progress passed all 9/9 linked boundary tests. Codes remain head_checks_passed with rule=declared; README documents this contract."
   "2026-10-16T19:35:17.666Z",codex-agent,"Round 1 npm acceptance: CI-identity-isolated npm run release:check passed 747/747 tests with zero failures/skips/cancellations/todos. Coverage is exact 100/100/100/100 across 13 authored modules: 10083/10083 lines and statements, 256/256 functions, 2858/2858 branches. All release gates passed at unchanged thresholds. CI-style pinned tracker health also passes; normal-home health reports local extension host version skew, with no tracked integrity/history drift. Bun release gate is running next."
   "2026-10-16T19:53:20.313Z",codex-agent,"Round 1 final acceptance: npm run release:check and bun run release:check both passed 747/747 tests with zero failures/skips/cancellations/todos using isolated CI-style Git identity. Both measured exact 100/100/100/100 coverage across 13 authored modules (10083 lines/statements, 256 functions, 2858 branches) at unchanged thresholds, with all static, privacy, audit, package, changelog and attestation gates passing. The fixed linked boundary suite passed 9/9; independent function revert proofs are recorded above. Declared-only success now explicitly states that repository required-check metadata was unavailable, without changing head_checks_passed or rule=declared. Caller declarations still require maintenance to match repository policy. Ready to push PR #140; item remains in progress for orchestrator verification."
+  "2026-10-16T20:39:26.384Z",codex-agent,"Round 2: replace whitespace splitting with one trimmed required context per line, dropping empty lines. Preserve the quoted newline-joined workflow argument, update CLI help and caller examples, and exercise exact names plus literal shell-shaped names through real workflow bash and SDK trackers. Record behavioural revert proof, then run unchanged npm and Bun release gates."
+  "2026-10-16T20:43:48.686Z",codex-agent,"Round 2 behavioural revert proof: restored only the required-context parsing stanza inside runCloseout to the pre-review repeated(...).flatMap(value => value.split(/\\s+/)) implementation, retaining all new tests and other changes. After npm run build, node --test --test-name-pattern=\"newline required contexts\" test/closeout-workflow.test.ts failed all 5/5 tests (0 passed; closed=0 instead of 1): single spaced name, spaces/parentheses/slashes, fleet example, empty lines/trimming, and literal shell-shaped names. Restored the line parser and rebuilt successfully. The fixed focused suite passed 18/18 with no skips or cancellations."
+  "2026-10-16T20:46:05.570Z",codex-agent,"The first pinned linked-test run hit its already-loaded default 120-second timeout before completion, with no assertion failures in completed cases. Set the new linked entry to timeout_seconds=1800 and rerun with an explicit --timeout 1800. This timeout is missing acceptance evidence, not a pass."
+  "2026-10-16T20:49:00.145Z",codex-agent,"Pinned node_modules/.bin/pm test pm-github-f1ev --run --progress --only-last --timeout 1800 passed the new linked command: 18/18 tests, zero failures/skips/cancellations/todos, no timeout. Test-result tracking is disabled by project policy, so this comment preserves the completed acceptance receipt without changing that policy."
+  "2026-10-16T21:13:33.378Z",codex-agent,"Round 2 npm run release:check passed with isolated CI-style Git identity: 754/754 tests, zero failures/skips/cancellations/todos. Exact 100/100/100/100 coverage across 13 authored modules: 10090/10090 lines and statements, 257/257 functions, 2866/2866 branches. All static, privacy, audit, package, changelog and attestation gates passed at unchanged thresholds. Tracker health passed with advisory telemetry/stale-item findings. Starting the Bun release gate on unchanged code."
+  "2026-10-16T21:41:47.426Z",codex-agent,"Round 2 final local acceptance: npm run release:check and bun run release:check both passed 754/754 tests with zero failures/skips/cancellations/todos using isolated CI-style Git identity. Both measured exact 100/100/100/100 coverage across 13 authored modules (10090 lines/statements, 257 functions, 2866 branches), at unchanged thresholds; all static, privacy, audit, package, changelog and attestation gates passed. The pinned linked suite passed 18/18 after replacing its default timeout with 1800 seconds. The 5/5 failing behavioural revert proof is recorded above. required-checks now preserves one trimmed literal context per line, drops blank lines, and retains safe quoted transport. README, CLI help and caller template show the new contract, including test (22) and test (26). No new PM CLI/SDK defect was identified. Caller declarations must still match repository policy. Ready to push PR #140; GitHub checks will be polled without thread replies or mentions. Item remains in progress for orchestrator verification."
 files[5]{path,scope}:
   .github/workflows/pm-github-closeout.yml,project
   closeout.ts,project
   index.ts,project
   test/closeout-workflow.test.ts,project
   test/closeout.test.ts,project
-tests[1]{command,scope,timeout_seconds,provenance{author,created_at,source_kind,source_ref}}:
-  "npm run build && npm run build:test && node --test --test-name-pattern=\"required-check metadata boundary|declared contexts|declared names\" test/closeout.test.ts test/closeout-workflow.test.ts",project,1800,codex-agent,"2026-10-16T16:10:42.723Z",local_mutation,fix/closeout-required-checks-unverified
+tests[2]:
+  - command: "npm run build && npm run build:test && node --test --test-name-pattern=\"required-check metadata boundary|declared contexts|declared names\" test/closeout.test.ts test/closeout-workflow.test.ts"
+    scope: project
+    timeout_seconds: 1800
+    provenance:
+      author: codex-agent
+      created_at: "2026-10-16T16:10:42.723Z"
+      source_kind: local_mutation
+      source_ref: fix/closeout-required-checks-unverified
+  - command: "npm run build && npm run build:test && node --test --test-name-pattern=\"newline|required-check metadata boundary|declared contexts|declared names|invalid selectors|closeout contract\" test/closeout.test.ts test/closeout-workflow.test.ts"
+    scope: project
+    timeout_seconds: 1800
+    provenance:
+      author: codex-agent
+      created_at: "2026-10-16T20:40:51.494Z"
+      source_kind: local_mutation
+      source_ref: fix/closeout-required-checks-unverified
+    updated_by: codex-agent
+    updated_at: "2026-10-16T20:44:12.196Z"
 docs[2]{path,scope}:
   docs/closeout-workflow-caller.yml,project
   README.md,project
diff --git a/.github/workflows/pm-github-closeout.yml b/.github/workflows/pm-github-closeout.yml
index 57d33a1d..852ae4e9 100644
--- a/.github/workflows/pm-github-closeout.yml
+++ b/.github/workflows/pm-github-closeout.yml
@@ -32,7 +32,7 @@ on:
         required: false
         default: "ci.yml"
       required-checks:
-        description: "Whitespace-separated required contexts to verify when protection metadata is unavailable"
+        description: "One required context name per line; spaces are preserved and blank lines ignored; used when protection metadata is unavailable"
         type: string
         required: false
         default: ""
@@ -290,6 +290,7 @@ jobs:
           # Full history catches up after missed schedules; the command skips terminal items.
           # Keep the report private until the proposed change has passed the privacy gate.
           required_args=()
+          # Forward the newline-joined list as one literal argument; TypeScript parses lines.
           if [[ -n "${REQUIRED_CHECKS:-}" ]]; then
             required_args+=(--required-checks "${REQUIRED_CHECKS}")
           fi
diff --git a/README.md b/README.md
index 5a6f9d66..b19f2caa 100644
--- a/README.md
+++ b/README.md
@@ -67,8 +67,8 @@ The check policy reflects the rules readable at invocation time.
 
 When either protection source is unreadable (HTTP 403/404), the row is ineligible
 with `required_checks_unverified`. Observed green runs cannot prove that every
-required context ran. Supply `--required-checks "CI CodeQL"` to declare a
-whitespace-separated set of required context names when metadata is unavailable.
+required context ran. Supply `--required-checks $'test (22)\ntest (26)'` (Bash) to declare a
+newline-separated set of required context names when metadata is unavailable.
 Every declared context must conclude **success** on the exact head (or have a
 successful latest legacy status); missing, pending, neutral or skipped contexts
 refuse closure. Readable requirements remain enforced, including app bindings.
@@ -556,9 +556,20 @@ inside the workspace and contain `package.json`. The SDK resolves the tracker
 from the repository root. The exact release version resolved from the workflow commit is installed and verified.
 
 `required-checks` is an optional workflow input with an empty default. Set it to
-all required context names (for example `"CI CodeQL"`) when the token cannot read
-branch-protection or ruleset metadata, as is common with `GITHUB_TOKEN`. The workflow
-passes this value as one quoted command argument. Without a declaration, unavailable
+all required context names, **one context name per line**, when the token cannot read
+branch-protection or ruleset metadata, as is common with `GITHUB_TOKEN`:
+
+```yaml
+with:
+  required-checks: |
+    test (22)
+    test (26)
+```
+
+Each line is trimmed and empty lines are ignored. Spaces within names such as
+`CI / build` are preserved; names are never split on spaces. The workflow passes
+the newline-joined list as one quoted command argument for TypeScript to parse.
+Without a declaration, unavailable
 metadata refuses closeout and prevents publishing a branch or PR. Keep this list
 aligned with the caller's required checks; it is a caller-owned policy declaration.
 
diff --git a/closeout.ts b/closeout.ts
index 53b178bf..7887f215 100644
--- a/closeout.ts
+++ b/closeout.ts
@@ -83,6 +83,14 @@ function repeated(options: Record<string, unknown>, key: string): string[] {
   return [...new Set((values as string[]).map(value => value.trim()))];
 }
 
+/** Parse one literal context per line, preserving spaces and ignoring blank lines. */
+function requiredContexts(value: unknown): string[] {
+  if (value === undefined) return [];
+  const values = Array.isArray(value) ? value : [value];
+  if (values.some(entry => typeof entry !== "string")) throw new Error("--required-checks requires newline-separated context names.");
+  return (values as string[]).flatMap(entry => entry.split("\n")).map(line => line.trim()).filter(Boolean);
+}
+
 /** Read one payload with the shared authenticated retry/redirect transport. */
 async function payload(api: CloseoutApi, endpoint: string): Promise<unknown> {
   return JSON.parse((await api.get(`${api.base}${endpoint}`)).body) as unknown;
@@ -358,8 +366,7 @@ export async function runCloseout(pmRoot: string, repo: string | undefined, opti
   });
   const since = options.since === undefined ? undefined : usage(() => isoDate(text(options.since)));
   if ((numbers.length === 0) === (since === undefined)) throw new CloseoutUsageError("Select either repeatable --pr <n> or --since <ISO date>.");
-  const declaredOptions = { "required-checks": options["required-checks"] ?? options.requiredChecks };
-  const declared = usage(() => repeated(declaredOptions, "required-checks")).flatMap(value => value.split(/\s+/));
+  const declared = usage(() => requiredContexts(options["required-checks"] ?? options.requiredChecks));
   const skip = new Set(usage(() => repeated(options, "skip")));
   const apply = options.apply === true && options["dry-run"] !== true && options.dryRun !== true;
   if (since !== undefined) {
diff --git a/docs/closeout-workflow-caller.yml b/docs/closeout-workflow-caller.yml
index f2eebcf9..395ade8c 100644
--- a/docs/closeout-workflow-caller.yml
+++ b/docs/closeout-workflow-caller.yml
@@ -23,5 +23,8 @@ jobs:
     with:
       npm-project-dir: "." # Use tools/oracle or another contained npm project when needed.
       post-closeout-script: "changelog:full" # Missing script skips; "" disables it.
-      required-checks: "" # Whitespace-separated context names (e.g. "CI CodeQL"); required when protection metadata cannot be read.
+      # One context per line; match your repository's required checks when metadata cannot be read.
+      required-checks: |
+        test (22)
+        test (26)
       ci-workflows: "ci.yml" # Whitespace-separated workflow filenames; "" disables dispatch.
diff --git a/index.ts b/index.ts
index 7c8405f7..6fdad97f 100644
--- a/index.ts
+++ b/index.ts
@@ -6688,7 +6688,7 @@ export default defineExtension({
         { long: "--repo", value_name: "owner/repo", description: "Repository (or PM_GITHUB_REPO)" },
         { long: "--pr", value_name: "n", repeatable: true, description: "Select a PR number (repeatable; conflicts with --since)" },
         { long: "--since", value_name: "ISO date", description: "Select PRs merged at or after this ISO date/time" },
-        { long: "--required-checks", value_name: "contexts", description: "Whitespace-separated required contexts when protection metadata is unavailable" },
+        { long: "--required-checks", value_name: "contexts", description: "One required context per line when protection metadata is unavailable (spaces preserved; blank lines ignored)" },
         { long: "--skip", value_name: "id", repeatable: true, description: "Never close this item (repeatable)" },
         { long: "--apply", description: "Close eligible items (default: dry-run)" },
         { long: "--dry-run", description: "Preview only; overrides --apply" },
diff --git a/test/closeout-workflow.test.ts b/test/closeout-workflow.test.ts
index bd8024e0..d52a41bc 100644
--- a/test/closeout-workflow.test.ts
+++ b/test/closeout-workflow.test.ts
@@ -155,7 +155,10 @@ esac
       head: { sha }, base: { ref: "trunk", repo: { full_name: "acme/widgets" } } });
     else if (url.includes("/protection/")) jsonResponse(res, 200, { contexts: ["CI"], checks: [] });
     else if (url.includes("/rules/branches/")) jsonResponse(res, 200, []);
-    else if (url.includes("/check-runs")) jsonResponse(res, 200, { total_count: 1, check_runs: [{ name: "CI", head_sha: sha, status: env.HEAD_STATUS ?? "completed", conclusion: env.HEAD_CONCLUSION ?? "success", app: { id: 7 } }] });
+    else if (url.includes("/check-runs")) {
+      const names = env.HEAD_NAMES ? JSON.parse(env.HEAD_NAMES) as string[] : ["CI"];
+      jsonResponse(res, 200, { total_count: names.length, check_runs: names.map(name => ({ name, head_sha: sha, status: env.HEAD_STATUS ?? "completed", conclusion: env.HEAD_CONCLUSION ?? "success", app: { id: 7 } })) });
+    }
     else if (url.includes("/statuses")) jsonResponse(res, 200, []);
     else jsonResponse(res, 404, {});
   });
@@ -210,7 +213,7 @@ test("closeout contract keeps sync safety helpers, exact pins, env bindings and
   assert.match(caller, /@RELEASE_COMMIT_SHA/);
   assert.match(caller, /^permissions: \{\}$/m);
   assert.match(caller, /gated-closeout:\n    permissions:\n      contents: write\n      pull-requests: write\n      checks: read\n      actions: write/);
-  assert.match(caller, /required-checks: ""/);
+  assert.match(caller, /required-checks: \|\n        test \(22\)\n        test \(26\)/);
   assert.match(workflow, /REQUIRED_CHECKS: \$\{\{ inputs.required-checks \}\}/);
   assert.doesNotMatch(caller, /issues: read/);
 });
@@ -403,11 +406,11 @@ test("compact PR summary is privacy gated before the bot branch is published", a
   assert.equal(existsSync(f.env.PR_STATE!), false);
 }));
 
-for (const mode of ["undeclared", "missing", "neutral", "skipped", "pending", "success", "metadata"]) {
+for (const mode of ["undeclared", "blank", "missing", "neutral", "skipped", "pending", "success", "metadata"]) {
   test(`required-check metadata boundary ${mode} uses the real workflow and command`, async () => fixture(async f => {
     f.env.PROTECTION_DENIED = mode === "metadata" ? "false" : "true";
     // A declared policy never replaces a complete readable policy.
-    f.env.REQUIRED_CHECKS = mode === "undeclared" ? "" : mode === "missing" || mode === "metadata" ? "CI missing" : "CI";
+    f.env.REQUIRED_CHECKS = mode === "undeclared" ? "" : mode === "blank" ? " \n\t\r\n " : mode === "missing" || mode === "metadata" ? "CI\nmissing" : "CI";
     if (["neutral", "skipped"].includes(mode)) f.env.HEAD_CONCLUSION = mode;
     if (mode === "pending") f.env.HEAD_STATUS = "queued";
     const before = readFileSync(join(f.pmRoot, "tasks", `${f.id}.toon`), "utf8");
@@ -423,8 +426,8 @@ for (const mode of ["undeclared", "missing", "neutral", "skipped", "pending", "s
       const item = (await get(f.id, {}, { pmRoot: f.pmRoot, noExtensions: true })).item;
       assert.match(String(item.actual_result), mode === "metadata" ? /required checks satisfied/ : /caller-declared checks concluded success; repository required-check metadata was unavailable; successful head checks: CI/);
     } else {
-      assert.equal(report.plan[0]!.code, mode === "undeclared" ? "required_checks_unverified" : "head_checks_failed");
-      if (mode === "undeclared") assert.equal(result.code, 1, result.out);
+      assert.equal(report.plan[0]!.code, mode === "undeclared" || mode === "blank" ? "required_checks_unverified" : "head_checks_failed");
+      if (mode === "undeclared" || mode === "blank") assert.equal(result.code, 1, result.out);
       assert.equal(readFileSync(join(f.pmRoot, "tasks", `${f.id}.toon`), "utf8"), before);
       assert.equal(existsSync(f.env.PR_STATE!), false);
       unpublished(f.git);
@@ -432,3 +435,25 @@ for (const mode of ["undeclared", "missing", "neutral", "skipped", "pending", "s
     }
   }));
 }
+
+for (const [scenario, names, input] of [
+  ["single name", ["CI / build"], "CI / build"],
+  ["spaces parentheses and slashes", ["CI / build (linux)", "CI / lint"], "CI / build (linux)\nCI / lint"],
+  ["fleet example", ["test (22)", "test (26)"], "test (22)\ntest (26)\n"],
+  ["empty lines and trimming", ["test (22)", "test (26)"], "\n  test (22) \r\n \t\n\t test (26)\t\n\n"],
+  ["literal shell names", ["$(id)", "CI / 'single' and \"double\" quotes", "$(touch injected)", "CI \\ * ? [ab]"], "$(id)\nCI / 'single' and \"double\" quotes\n$(touch injected)\nCI \\ * ? [ab]"],
+] as const) {
+  test(`newline required contexts ${scenario} survive the real workflow and command`, async () => fixture(async f => {
+    f.env.PROTECTION_DENIED = "true";
+    f.env.HEAD_NAMES = JSON.stringify(names);
+    f.env.REQUIRED_CHECKS = input;
+    const result = await f.run([...setup, "Close delivered tracker items"]);
+    assert.equal(result.code, 0, result.out);
+    const report = JSON.parse(readFileSync(join(f.base, "pm-github-closeout.json"), "utf8")) as { closed: number; checks: { rule: string; passed: string[] }[] };
+    assert.equal(report.closed, 1);
+    assert.equal(report.checks[0]!.rule, "declared");
+    assert.deepEqual([...report.checks[0]!.passed].sort(), [...names].sort());
+    assert.equal((await get(f.id, {}, { pmRoot: f.pmRoot, noExtensions: true })).item.status, "closed");
+    assert.equal(existsSync(join(f.root, "injected")), false);
+  }));
+}
diff --git a/test/closeout.test.ts b/test/closeout.test.ts
index 65771419..66460456 100644
--- a/test/closeout.test.ts
+++ b/test/closeout.test.ts
@@ -369,7 +369,7 @@ test("invalid selectors fail before HTTP or writes", async () => tracker(async (
   await withMockGithub(fixture(null), async server => {
     for (const options of [{}, { pr: [] }, { pr: false }, { pr: "" }, { pr: ["42", 43] }, { pr: "0" }, { pr: "1.2" }, { pr: "9007199254740992" },
       { pr: "42", since: "2026-10-01" }, { since: "yesterday" }, { since: "2026-02-30" }, { since: "2026-10-01T12:00:00" },
-      { since: "2026-10-01T99:00:00Z" }, { since: false }, { since: "" }, { pr: "42", skip: [] }, { pr: "42", skip: true }, { pr: "42", "required-checks": "  " }, { pr: "42", "required-checks": false }]) {
+      { since: "2026-10-01T99:00:00Z" }, { since: false }, { since: "" }, { pr: "42", skip: [] }, { pr: "42", skip: true }, { pr: "42", "required-checks": ["CI", 42] }, { pr: "42", "required-checks": false }]) {
       await assert.rejects(() => closeout(root, options), CloseoutUsageError);
     }
     for (const repository of [undefined, "invalid", "bad/repo/name", "./repo", "../repo", "fixture/.", "fixture/.."]) await assert.rejects(() => runCloseout(root, repository, { pr: "42" }, api()), CloseoutUsageError);
@@ -738,13 +738,13 @@ test("declared contexts verify exact names and latest statuses without requiring
   }), async () => {
     const before = snapshot(root);
     assert.equal((await checkCloseoutHead(api(), repo, pull(link(ids[0]!)))).code, "required_checks_unverified");
-    for (const names of ["CI Legacy", "CI missing"]) {
+    for (const names of ["CI\nLegacy", "CI\nmissing"]) {
       const report = await closeout(root, { pr: "42", apply: true, "required-checks": names });
       assert.equal(report.closed, 0);
       assert.equal(report.plan[0]!.code, "head_checks_failed");
       assert.equal(snapshot(root), before);
     }
-    const report = await closeout(root, { pr: "42", apply: true, "required-checks": "  CI\tlegacy\nCI  " });
+    const report = await closeout(root, { pr: "42", apply: true, "required-checks": "  CI\nlegacy\nCI  " });
     assert.equal(report.closed, 1);
     assert.equal(report.checks[0]!.rule, "declared");
     assert.equal(report.checks[0]!.code, "head_checks_passed");
@@ -765,3 +765,23 @@ test("declared names cannot bypass a readable app binding", async () => tracker(
     assert.equal(snapshot(root), before);
   });
 }));
+
+test("newline context arrays preserve literal names and empty declarations fail closed", async () => tracker(async (root, ids) => {
+  await withMockGithub(fixture(link(ids[0]!), (request, response) => {
+    if (request.url!.includes("/protection/") || request.url!.includes("/rules/branches/")) jsonResponse(response, 403, {});
+    if (request.url!.includes("/check-runs")) jsonResponse(response, 200, { total_count: 2, check_runs: [run({ name: "test (22)" }), run({ name: "test (26)" })] });
+  }), async () => {
+    const before = snapshot(root);
+    for (const names of [[], "", " \n\t\r\n ", ["", "\n"]]) {
+      const report = await closeout(root, { pr: "42", apply: true, "required-checks": names });
+      assert.equal(report.plan[0]!.code, "required_checks_unverified");
+      assert.equal(snapshot(root), before);
+    }
+    // A former space-separated list is one literal name, so observed runs cannot satisfy it.
+    const missing = await closeout(root, { pr: "42", apply: true, "required-checks": "test (22) test (26)" });
+    assert.equal(missing.plan[0]!.code, "head_checks_failed");
+    assert.equal(snapshot(root), before);
+    const report = await closeout(root, { pr: "42", apply: true, requiredChecks: ["  test (22)\n\n", "test (26)\r\n"] });
+    assert.equal(report.closed, 1);
+  });
+}));