Repository navigation
Harden mutation provenance and crawler failmaps - #789
Merged
Merged
Conversation
d6g8k5htny-coder
marked this pull request as ready for review
November 9, 2026 22:19
Codex Review Summary: ✅ CompletedThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem and change
The inventory, freshness and reference-crawler mutation graders could credit failures without establishing the intended oracle: equal-but-distinct code objects, lifecycle-invoked selected bodies, a spoofed unittest filename, or a runtime error relabeled as a failure. The crawler also omitted its parent's optimized mode and accepted skipped or expected-failure cases. Actual adverse controls and prior failures are retained.
The candidate uses live code-object identity, execution-phase tracking and actual assertion type, restores the inventory status-isolation failure/count, refuses a skipped selected target, and supports the tested standard-library assertion forms. The crawler checks the full selected population, child mode and non-skipped outcomes. Its v3 failmaps bind crawler and grader bytes, Python version and optimization mode; v2 maps are not rebound. Serialized failmaps do not authenticate an untrusted producer, and arbitrary test-class helper origins remain unsupported.
The candidate registers the prepared Node runner, records the separate Math #559 grader defect, and updates the five canonical handoff surfaces and decisions. It preserves earlier records in Git and appends corrections to their attribution and baseline label. It changes no mathematical statement, bound, axiom, Lean outcome, quarantine or scientific disposition.
Observed checks and source boundaries
08aabcbdd2a937dfbff51d230166cd19d4149684and local commitd44af03cf48a655f50948cb751b1813b1ee73ab0have identical treeea3c2908b515fb04042100cd430de2da81efb54a, based on Main1902aac113761974c4569f78acb6a8baadabb7f1. Its latest follow-up changes only the six canonical documentation/decision files; the branch update used the expected head and no force.bd23db4e9998c30e2bbfee912cb58ae7f150f822, tested merge8b4c8656bf35bdccfb9decddef29194ff68e6289, crawler run42705939370completed 50 frozen and 23 grader tests per mode, with 16 grader mutants per mode and 109/109 designated source mutations in each mode. All nine artifact ZIPs matched their declared hashes and passed CRC; merged population, source hashes and failure provenance were checked. The prepared Node run42705939654executed 33 helper controls and 767 Node tests per mode with no skips, failures, cancellations or todo cases; its 39 fixture bytes per mode and TAP digests were checked.de9db13d21b53ab33dda09818e01d53979d90f8ahas the candidate tree. Crawler run42707612430produced 109/109 designated source mutations in each mode; its nine original artifacts, source hashes, selected population and recorded provenance were checked. Node run42707612474produced 767/767 tests and 33 helper controls per mode with no skips; all 39 fixture bytes per mode, TAP and source identities were checked. The original and filtered evidence is retained with SHA-256462b9209a7a9cf720d68918544df1ec4f993d5dc499e5840d3cebb7a383bec60. Normal integration remains pending; these are component observations. Workspace Review separating-cap extension: fixed-axis GLOBAL elder-defect O(r^3), beyond local paths #58 pins the preceding Main and needs a fresh guarded comparison after any Main integration. The earlier wrong-order population rejection, overlapping-guard mutation miss, self-mutation timeout, and 766/767 Node failure remain preserved.No final G0–G11 pass, whole-federation audit, no-skip full regression, CBZ-0, terminal Partial or independent scientific acceptance is claimed.