Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
22 commits
Select commit Hold shift + click to select a range
1a8989f
Merge commit '5aa6fbac2c66fac0e6e2332b7e95cacc3c37350e' into HEAD
codex Oct 29, 2026
c32433a
Compose strict formal rejection audit with producer-attempt binding
codex Oct 29, 2026
c964f87
Require complete reason-bound compiled-auditor mutation kills
codex Oct 29, 2026
5494cfe
Retain exact source-bound receipt composition verification evidence
codex Oct 29, 2026
5d25cad
Bind formal receipts to immutable owned source before and after execu…
codex Oct 29, 2026
ce4277c
Retain successor receipt-composition evidence and compatible source f…
codex Oct 29, 2026
aa0d59a
Merge commit '22567eeff95f3184048b3e2190ecab908bfc115a' into HEAD
codex Oct 29, 2026
c7b8161
Retain exact custody failure and final verified producer evidence
codex Oct 29, 2026
7b16577
Reconcile landed receipt-attempt composition with stronger source cus…
codex Oct 29, 2026
f6e93a1
fix(formal): bind mutation kills to executed test bodies
d6g8k5htny-coder Oct 29, 2026
65c07f8
docs(formal): preserve observed hosted qualification and recovery rec…
d6g8k5htny-coder Oct 29, 2026
d279c76
Compare cited Git object identities in node-status freshness checks
d6g8k5htny-coder Oct 29, 2026
5d35ef0
Preserve newly landed kernel evidence in freshness comparator composi…
d6g8k5htny-coder Oct 29, 2026
75d91e7
Bind comment freshness to its canonical parent and reject duplicate J…
d6g8k5htny-coder Oct 29, 2026
790bcf9
docs(cbz): register finite MomentTail evidence and reproducible checks
d6g8k5htny-coder Oct 30, 2026
9c89e1c
Merge branch 'main' into codex/cbz-empirical-registration-01a11d37-20…
d6g8k5htny-coder Oct 31, 2026
5d3db3c
Merge main (56279583) into #706; take main's .gitattributes
claude Oct 31, 2026
e5f8c05
ci(node-status): run the freshness suite and its mutation controls on…
claude Oct 31, 2026
37ba7c6
Merge main (56279583) into #709 after applying the reviewed workflow …
claude Oct 31, 2026
7c75c1e
Integrate #717 (9c89e1c8)
claude Oct 31, 2026
58b5592
Integrate #706 (5d3db3c6)
claude Oct 31, 2026
545cf9e
Integrate #709 (37ba7c67)
claude Oct 31, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 9 additions & 2 deletions .github/workflows/current-site.yml
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,8 @@ name: Current research site
- 'tests/test_current_site*'
- 'tools/node_status_refresh.py'
- 'tests/test_node_status_refresh.py'
- 'tests/test_node_status_freshness.py'
- 'cbz/evidence/node-status-freshness/run_mutations.py'
- 'tests/fixtures/node_status_refresh/**'
- '.github/workflows/current-site.yml'
schedule:
Expand Down Expand Up @@ -56,8 +58,13 @@ jobs:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1
with:
persist-credentials: false
- name: Offline freshness tests
run: python3 -B -m unittest tests.test_node_status_refresh
- name: Offline freshness tests and mutation controls
run: |
python3 -B -S -m unittest tests.test_node_status_freshness tests.test_node_status_refresh -v
python3 -B -O -S -m unittest tests.test_node_status_freshness tests.test_node_status_refresh -v
python3 -B -S cbz/evidence/node-status-freshness/run_mutations.py > "$RUNNER_TEMP/node-status-mutations.json"
# Pull requests run only the offline steps above. The live refresh reads the GitHub API,
# so it runs on main pushes, schedules and manual dispatch, never on a pull request.
- name: Report-only freshness refresh (read-only GitHub API)
if: github.event_name != 'pull_request'
env:
Expand Down
4 changes: 2 additions & 2 deletions .github/workflows/formal-verification.yml
Original file line number Diff line number Diff line change
Expand Up @@ -73,8 +73,8 @@ jobs:
"$HOME/.elan/bin/elan" toolchain install "$(cat formal/lean-toolchain)"
- name: Gate controls in both interpreter modes
run: |
python -B -S -m unittest tests.test_formal_gate tests.test_formal_negative_diagnostics tests.test_formal_compiled_inventory -v 2>&1 | tee /tmp/tests-normal.log
python -B -O -S -m unittest tests.test_formal_gate tests.test_formal_negative_diagnostics tests.test_formal_compiled_inventory -v 2>&1 | tee /tmp/tests-optimized.log
python -B -S -m unittest tests.test_formal_gate tests.test_formal_negative_diagnostics tests.test_formal_compiled_inventory tests.test_formal_source_custody -v 2>&1 | tee /tmp/tests-normal.log
python -B -O -S -m unittest tests.test_formal_gate tests.test_formal_negative_diagnostics tests.test_formal_compiled_inventory tests.test_formal_source_custody -v 2>&1 | tee /tmp/tests-optimized.log
- name: Kernel lane — build, leanchecker, axiom audit, negative controls, receipt
env:
GITHUB_REPOSITORY: ${{ github.repository }}
Expand Down
7 changes: 7 additions & 0 deletions DECISIONS.md
Original file line number Diff line number Diff line change
Expand Up @@ -27,3 +27,10 @@ Append-only. Historical source outcomes remain in their original records.
[2026-10-30 00:24:29 UTC] [Phase 3] [Layer C] Restore AGENTS.md canonical visible label “Agent Message Board #619” without changing its URL or any routing test. Actual OpenAI/Codex chat01a0bbb5 diagnosed final-head1296 navigation run40978732830/job123813126343: 1693 tests, one routing assertion failure, 37 skips, exit1. The documentation rewrite shortened the label to #619, so the existing routing contract found zero canonical links. Raw failed artifact12944475914 (43651 bytes, SHA25659b8b4cee6d3a676c91be99548bb588006373b0e57f6a998f794cc9f0cc6940a) is preserved. Exact original routing module reproduced 1 failure among23 tests in normal and optimized local Python3.9.6; correcting only the visible label passed all23 in each mode. No test/predicate or Layer A invariant changed. Current-source hosted verification remains required; previous P15/replay successes retain their actual1296 source. Reversible: source ancestry and failed output retained.

[2026-10-30 00:36:25 UTC] [Phase 3] [Layer C/A8] Add a separate strict recorded-mutation qualifier after the frozen replay runner; preserve all four historical Python sources and SHA256SUMS. Actual OpenAI/Codex navigation reviewer demonstrated six synthetic wrong-cause/mixed-error false credits in the old grader, with zero real child executions (713/6511866871). Actual six hosted outcomes remain valid on inspection. OpenAI/Codex nav679 implemented two new files;14 tests passed normal/optimized locally, six source mutants killed in both modes, and separate navigation review repeated parser/CLI and authentic/adverse checks. Parent chat01a0bbb5 composes the two files, mandatory CI post-check/count guards and this registration. Qualification is source/report/raw-log consistency, never unsigned execution authentication or proof/global-tool admission. The existing foundation A8 grader pair is a distinct unresolved source issue (695/6511848166). Reversible: frozen sources, original reports, negative probes and all failed attempts retained. Final changed-input hosted qualification is required.

| UTC | Phase | Layer | Action and rationale | Reversible | Evidence |
|---|---|---|---|---|---|
| 2026-10-30 00:36:59 | 2 | B/tooling | Registered the finite MomentTail candidate Math545/e1cf617b and native self-test scope/commands under GOVERNANCE and STATUS. Actual hosted40980378198/attempt1/job123818653248 at tested mergef38124cb/tree87cd19c6 passed all6 commands,20+12 tests per mode and10019distinct cases per mode; artifact12944860659 ZIP SHA256dcf13fa0124c51091731786aec81c03a32b5fe53a8e49f689b1e5901c17f415d verified. General coverage, global CLOSED enforcement and finalG0-G11 remain unmet. ExecutorOpenAI/Codex chat01a11d37 with delegated authors/reviewers, no scientific credit. | Yes | [Hosted readback](https://github.com/d6g8k5htny-coder/Math-/pull/545#issuecomment-6511948766); exact contract and component source |
| 2026-10-30 00:36:59 | 3 | C/runtime | Clarified Atlas control runtime in its existing guide: CPython3.9.6 yields9 preserved assertion/subtest failures across2 grader methods because callbacks arrive after teardown; currentMain11e288 hosted3.12.15 run40976675758/attempt1/job123806316839 passed31controls and2required browser methods in each mode. Preserve failures and historical30/45 counts; update the old exclusive-writer guidance to current fluid roles. No grader, test, workflow or browser policy changed. | Yes | [Actual hosted run](https://github.com/d6g8k5htny-coder/main/actions/runs/40976675758); exact source/decoded-log diagnosis; docs/cbz/ATLAS_BROWSER_REGRESSIONS.md |
| 2026-10-30 00:36:59 | 2 | C/composition | Ruling: prepare this four-document contribution as an isolated successor of713/62292ad6 and a stacked PR, without changing the existing713 branch or resetting its current checks. Retarget to Main after that base actually lands and recheck the resulting composition. Source guards prevent overwrites; the pending base is an integration dependency, not an exclusive coordinator lock. | Yes | Original713 source and parent; current team Master; guarded source/readback records |
| 2026-10-30 00:40:56 | 2 | C/source | Guard observed713 advance62292ad6 toaf3996f5. Recompose this same bounded registration on freshaf3996f5, preserving its strict replay-qualifier GOVERNANCE section and complete prior ledger. Original author corrected review finding: GOV and STATUS now print all6 actual reproduction commands, including unchanged MomentTail unittest discovery normal/optimized, rather than only4. Private prior proposal retained; no old result rebound to new code. | Yes | Current713 source; bounded documentation review; exact command-to-native-CI comparison |
56 changes: 56 additions & 0 deletions GOVERNANCE.md
Original file line number Diff line number Diff line change
Expand Up @@ -122,6 +122,62 @@ Mission tools must be finished, documented with exact usage and verification,
and registered here and in `TOOLING.md` before final admission. Tools are a
means to closure, never a reason to open a general infrastructure project.

### Finite MomentTail empirical component: bounded registration

**Registered 2026-10-30 00:36:59 UTC; global admission remains pending.**
[Math PR #545](https://github.com/d6g8k5htny-coder/Math-/pull/545) supplies
`cbz/finite_moment_tail_20261030/adapter.py`, its self-tests and complete
[CONTRACT.json](https://github.com/d6g8k5htny-coder/Math-/blob/e1cf617b6bc979b75d420f28df17ddad59426522/cbz/finite_moment_tail_20261030/CONTRACT.json)
at head `e1cf617b6bc979b75d420f28df17ddad59426522`, tree
`87cd19c612906a61bca7659fa24e50a3958e0eb3`. This registry entry does not
represent that candidate as merged or update the federation's existing pins.

Purpose: empirical falsification of the finite rational specialization of
`ResearchFormalCoreR1.p02_lm009_moment40_tail`. It uses exact Fraction/integer
decisions for normalized finite laws, signed jet values, any positive r and
epsilon, an actually sufficient fortieth-moment bound, and the strict event.
Actual Lean/test byte hashes are checked before evaluation. Invalid inputs,
false premises, unsupported targets, altered source and empty coverage refuse
success. The two source-file checks do not attest dependencies or run Lean.

From an isolated checkout of that exact Math source, run:

```
python3 -B cbz/finite_moment_tail_20261030/test_adapter.py -v
python3 -B -O cbz/finite_moment_tail_20261030/test_adapter.py -v
python3 -B -m unittest discover -s formal/tests -p test_moment_tail.py -v
python3 -B -O -m unittest discover -s formal/tests -p test_moment_tail.py -v
python3 -B cbz/finite_moment_tail_20261030/adapter.py --source-root . campaign --seed 20261030 --count 10019
python3 -B -O cbz/finite_moment_tail_20261030/adapter.py --source-root . campaign --seed 20261030 --count 10019
```

Self-verification: [run 40980378198, attempt 1](https://github.com/d6g8k5htny-coder/Math-/actions/runs/40980378198),
job `123818653248`, actually ran CPython 3.11.16 on merge
`f38124cbef8f8950dcc59fe0e0155d45cd2b767b` with the same tree. Component
20/20 and original MomentTail 12/12 tests passed in each mode; all six documented
commands exited zero. Each campaign observed the same 10,019 distinct valid
cases with zero violations/refusals. Five implementation mutants produced 20
intended child assertion failures with zero errors across parent/child modes:
three test wrong bounds on valid premises, one tests strict-event output, and
one tests false-premise admission. The latter two are separate semantic/admission
controls, not mathematical bound counterexamples. Artifact `12944860659` was
whole-byte verified against SHA256
`dcf13fa0124c51091731786aec81c03a32b5fe53a8e49f689b1e5901c17f415d`;
its actual source hashes, outputs, counters and child identities were read back.

Gate mapping: G0/G5/G6 use the documented self-tests, unchanged regressions and
adverse controls; the native `finite-moment-tail-self-test` job runs on every PR,
main push and manual dispatch and retains failed outcomes. G8/G9/G11 require
this exact scope, runnable commands and self-verification record. G2-G4 retain
the full axiom/build/kernel requirements, and G7 still requires federation
pins and contextual AST/type/theorem synchronization. This component neither
installs a global CLOSED guard nor supplies the general-real, Gaussian,
weighted/family or remaining unsupported coverage. The full empirical harness,
drift tool admission and common final G0-G11 run remain unfinished.
Maintenance owner: none; self-contained. Actual executor: OpenAI/Codex chat
`01a11d37` with delegated implementation and nonauthor engineering review;
same provider, scientific effect NONE.

## Team structure and coordination

**B-007 (owner expansion dated 2026-10-29):** Operate as a fluid team, not a
Expand Down
21 changes: 21 additions & 0 deletions STATUS.md
Original file line number Diff line number Diff line change
Expand Up @@ -45,6 +45,27 @@ not full AST/type/theorem synchronization. New gating tools require documented
commands and self-verification before admission; unavailable or unsupported
coverage remains explicit. No documented command here runs all G0-G11 yet.

The finite MomentTail component is a registered bounded candidate in
[Math PR #545](https://github.com/d6g8k5htny-coder/Math-/pull/545), exact source
`e1cf617b6bc979b75d420f28df17ddad59426522`. In an isolated checkout of that Math
source, run all six commands:

```
python3 -B cbz/finite_moment_tail_20261030/test_adapter.py -v
python3 -B -O cbz/finite_moment_tail_20261030/test_adapter.py -v
python3 -B -m unittest discover -s formal/tests -p test_moment_tail.py -v
python3 -B -O -m unittest discover -s formal/tests -p test_moment_tail.py -v
python3 -B cbz/finite_moment_tail_20261030/adapter.py --source-root . campaign --seed 20261030 --count 10019
python3 -B -O cbz/finite_moment_tail_20261030/adapter.py --source-root . campaign --seed 20261030 --count 10019
```

The [contract](https://github.com/d6g8k5htny-coder/Math-/blob/e1cf617b6bc979b75d420f28df17ddad59426522/cbz/finite_moment_tail_20261030/CONTRACT.json)
and [bounded governance registration](GOVERNANCE.md#finite-momenttail-empirical-component-bounded-registration)
record the six-command hosted job, exact source/test evidence and limits.
Its CPython 3.11.16 run actually passed; it covers one finite specialization,
not a federation-wide empirical gate or a new CLOSED decision. Existing pins
and the global unsupported inventory are unchanged by this registration.

## Team and reconstitution

Use the discovery, proof-closure, tooling, verification, integration and
Expand Down
32 changes: 32 additions & 0 deletions cbz/evidence/formal-lifecycle-db642-20261029/COMPOSITION.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,32 @@
{
"observed_at": "2026-10-29T20:38:12.909462+00:00",
"reviewed_head": "d676b961bc016110fa990de549a3d07888d5af24",
"native_701_base": "7b1657700aca5f8e9c391513c7974009688945ab",
"native_main": "c93cec70223281492709ff74ca84fed5dc6e40e8",
"composed_local_head": "658d6bc5b5a5df0cc3a50fdd4a8443f547d8e21b",
"composed_tree": "dde93df679476bedda9a833d853a0dbe26da5cf1",
"composition_changed_paths": [
"docs/RECON_NOVELTY_20260925.md"
],
"reviewed_source_bytes_unchanged": true,
"identities": {
"tests/test_formal_compiled_inventory.py": {
"git_blob": "91aa1d8c8cc862995bb66d40e2c7788d6fce46cc",
"sha256": "1e171f45a75b88ded8dc5d758edc835841546080489b01c38b38b876c7896fea"
},
"tests/test_formal_source_custody.py": {
"git_blob": "abc00f10bab4e198dcd95d28e1cfeb8994057fec",
"sha256": "6a45d063643362939ae703c7ba594c160e480f61dc0d797fbabf287f19850313"
},
"formal/manifest.json": {
"git_blob": "71097be079ed5320ab5bd404e569973328ab9644",
"sha256": "77679ec1514124f7c144a113f63514202029b53361dd5e4102ab07d8670a35f7"
},
"cbz/evidence/formal-lifecycle-db642-20261029/verify_classifier_mutations.py": {
"git_blob": "b163ac560f18104f344d2a4c3401d81a88ea618e",
"sha256": "a40a282624e67d4682b79f9b91577286c68063127ad7dea770c11c030531339f"
}
},
"tests_replayed_for_doc_only_composition": false,
"scope": "Git tree/source identity check only; not kernel, runtime or federation qualification"
}
5 changes: 5 additions & 0 deletions cbz/evidence/formal-lifecycle-db642-20261029/DECISIONS.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,5 @@
[2026-10-29T20:07:12.604016+00:00] [phase2] [layer C] Implement a bounded additive Main701 lifecycle harness amendment after exact current PR/claim checks found no separate matching correction pickup. Rationale: closes the demonstrated A8 false-kill contract while preserving existing producer-custody ownership and B-003 integration. Reversible: yes. Evidence:701comment6506371965, pickup6507053979; initial base7b165770. This is not a lease-expiry transfer.

[2026-10-29T20:38:12.909462+00:00] [Phase2] [LayerC] Source review and composition: task plus whole-branch reviews found no actionable issue ond676b961bc016110fa990de549a3d07888d5af24; current Mainc93cec70223281492709ff74ca84fed5dc6e40e8 combines with only its published novelty-memo delta. Exact reviewed sources remain unchanged. Additive delivery uses native701 parent and currentMain second parent; existing owners/claims unchanged. Reversible commits. Evidence: COMPOSITION.json, task-1-review.md, whole-branch-review.md.

[2026-10-29T21:05:15.927786+00:00] [Phase3/4] [LayerC] Actual hosted formal qualification and observed native check states atf6 preserved in completion/. Runtime observed3.11.17; historical3.12 local skips remain unchanged; queued checks stay queued. One evidence-only supplement preserves Math own-local recovery and exact prior-source results; later commit checks remain separate. No proof/test/manifest/runtime/workflow byte changed.
Loading
Loading