diff --git a/docs/analysis/full-suite-reds-ledger-2026-07.md b/docs/analysis/full-suite-reds-ledger-2026-07.md index 12b92c2e..1cfac10e 100644 --- a/docs/analysis/full-suite-reds-ledger-2026-07.md +++ b/docs/analysis/full-suite-reds-ledger-2026-07.md @@ -10,7 +10,7 @@ Base: `forgejo/main` @ `54228d45` |------|---------|--------| | Fast | `pytest -m "not quarantine and not slow" -n auto -q` | **5 failed**, 11660 passed, 106 skipped (pre-fix) | | Fast recheck | same after fix | (see verification below) | -| Slow | `pytest -m "slow and not quarantine" -n auto -q` | running / recorded at PR time | +| Slow | `pytest -m "slow and not quarantine" -n auto -q` | **1 failed, 909 passed, 1 skipped, 1 xfailed** (37m06s, 2026-07-15 sweep @ `8f67590d`+#47). The 1 red = `test_public_showcase.py::TestRuntimeBudget` — stale constant pin (30) vs the deliberate 640dbe8f re-budget (60); fixed alongside this row. Slow lane is otherwise clean → the historical "~31 reds" figure is fully retired. | | Full | `pytest -m "not quarantine"` | union of fast + slow | The brief’s “~31 reds” appears to reflect an older main tip. On current `forgejo/main`, the **non-quarantine fast lane** surfaces **exactly 5** dedicated failures. Smoke/pack/lane gates stay green because they never execute these nodes. diff --git a/tests/test_adr_0241_governance_p12.py b/tests/test_adr_0241_governance_p12.py index a295282b..5d9cbb38 100644 --- a/tests/test_adr_0241_governance_p12.py +++ b/tests/test_adr_0241_governance_p12.py @@ -4,7 +4,8 @@ Does not re-execute the full physics suite. Asserts the load-bearing governance surfaces for ADR-0241/0242 cohesion remain present and honest: * runtime_contracts documents off-serve quarantine + epistemic standing * acceptance checklist maps C0–C8 to tests - * ADRs stay Proposed (ready for Joshua) — not self-Accepted + * ADR statuses carry recorded ruling provenance (ratified 2026-07-15; + a silent status flip in either direction fails) * cohesion suite still names I-01…I-05 pins """ @@ -82,21 +83,37 @@ def test_cohesion_suite_names_entity_invariants(): assert name in text, f"missing entity pin {name}" -def test_adrs_ready_for_acceptance_not_self_accepted(): - """P12: implementation complete → Proposed + ready; Joshua alone Accepts.""" +def test_adrs_accepted_with_recorded_ruling_provenance(): + """P12 (post-ratification): status flips are valid ONLY with provenance. + + Joshua Shay ruled "Ratify" on 2026-07-15 (D10 acceptance packet §8), so + the anti-self-Accept guard evolves rather than dies: each ADR must say + Accepted WITH the ratification provenance inline, and the packet must + carry the ruling record. A silent status flip in either direction — an + Accept without provenance, or a quiet demotion — fails here. + """ + packet = ( + _ROOT / "docs/audit/adr-0241-0242-acceptance-packet-2026-07-15.md" + ).read_text(encoding="utf-8") + assert "## 8. RULING RECORD" in packet, "ruling record section missing" + assert "RATIFIED — Joshua Shay, 2026-07-15" in packet, "ruling attribution missing" + for rel in ( "docs/adr/ADR-0241-wave-field-driven-hyperbolic-atlas-and-resonant-cognition.md", "docs/adr/ADR-0242-atlas-packing-and-fibonacci.md", ): text = (_ROOT / rel).read_text(encoding="utf-8") - # First status line must remain Proposed until human Accept. status_line = next( (ln for ln in text.splitlines() if ln.startswith("**Status**")), "", ) - assert "Proposed" in status_line, f"{rel} lost Proposed status" - assert "Accepted" not in status_line, f"{rel} must not self-Accept" - assert "Joshua" in status_line or "ready" in status_line.lower() + assert "Accepted" in status_line, f"{rel} lost Accepted status" + assert "ratified by Joshua Shay" in status_line, ( + f"{rel} status lacks ratification provenance" + ) + assert "acceptance-packet" in status_line, ( + f"{rel} status must cite the ruling packet" + ) def test_serve_quarantine_list_matches_cohesion_ast_pin(): diff --git a/tests/test_public_showcase.py b/tests/test_public_showcase.py index 2171dc04..962bc007 100644 --- a/tests/test_public_showcase.py +++ b/tests/test_public_showcase.py @@ -162,5 +162,11 @@ class TestPureCompositionGate: class TestRuntimeBudget: - def test_budget_is_thirty_seconds(self) -> None: - assert MAX_RUNTIME_SECONDS == 30 + def test_budget_is_sixty_seconds(self) -> None: + # 30 → 60 was a deliberate re-budget (commit 640dbe8f: post-CGA + # substrate work makes a cold RegisterTour alone ~30s+; see + # evals/public_demo/contract.md "Known Environment Caveat"). This slow + # pin was missed then because the slow lane hadn't been run since + # (nightly lives on the billing-locked GitHub side) — caught by the + # 2026-07 slow-lane sweep. Pin the deliberate value. + assert MAX_RUNTIME_SECONDS == 60