Repository navigation
Research audit: PR741 binding accepts omitted replay evidence (no new kappa) - #838
Merged
Mister-Meeseeks merged 2 commits intoOct 13, 2026
Conversation
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The pinned PR741 composition checker can return PASS after a required stage is replaced at unchanged stage count, or after an artifact and its receipt entry are both removed. This research audit reproduces 111 accepted evidence mutations against actual temporary files, including public-CLI witnesses with the phase stage replaced and with 42 files deleted. It adds an optional coverage predicate that rejects all 117 mutations while retaining the original checker and accepting pristine evidence.
All additions are confined to
research/frontier741-receipt-coverage-audit/: exact outcomes, source/Git-blob pins, a standard-library verifier, local scope claim and a scoped research note. Base:7ee02b7c96d85f22b38d28bd7ed49bb8c1aeb261; assignment seed 72, adversarial audit. The selected frontier advanced from PR741 to PR799 during the work. The audited checker and all nine runtime inputs remain unchanged at the refreshedf567884aeb8851b6e794af1184119a3bfbe1e631; the moving selection metadata is historical context rather than a runtime dependency. PR799 calls the audited checker for ancestry, but this contribution does not audit its full wrapper or new construction.Reproduce from the repository root:
Expected:
PASS_SCOPED_PR741_RECEIPT_COVERAGE_AUDIT 111 altered-evidence cases accepted by original; 117 mutations rejected by supplement; pristine accepted.The replay passes with normal and optimized Python and from a relocated input package; altered checker source and expected evidence are rejected. Seven relevant existing regression tests, strict type checking and the unchanged selected-record check pass. Fresh supplier reconstruction, the full repository suite and remote CI were not rerun or claimed. The published checker, source packages, selected records and
upstream/are unchanged.This is an evidence-admission finding, not a counterexample to the construction or multiplication theorem, and no new kappa is claimed. The inherited all-size/compiler, common-label/mutable-source, hosting, precision/tape and analytic obligations remain. The supplemental predicate does not authenticate arbitrary logs or re-prove the certificates.
Credit Douglas Colkitt's binding and retained replay, Chafik Boukhalfa's PR741 construction, DreamingOfClouds' PR589 transport on eumemic's PR555 supplier, and the retained predecessor notices. PR704's ledger audit, PR740's finite-menu controls and PR777's frame checker have distinct scopes. Contribution by Lucas (@ljste), with substantial OpenAI Codex assistance, under Apache-2.0; no independent human or delegated-agent review claimed. No pinned coordination issue existed, so the claim was recorded locally.