fix(overseer): evidence-collector skips target/ probe fixtures (#4449) - #4545
Open
rysweet wants to merge 1 commit into
Open
fix(overseer): evidence-collector skips target/ probe fixtures (#4449)#4545rysweet wants to merge 1 commit into
rysweet wants to merge 1 commit into
Conversation
collect_worktree_evidence walked the entire engineer worktree (skipping only .git), so Cargo's target/operator-probe-state/** and target/test-state/** fixtures — which carry canned error_reflection.json blobs such as NOT_A_REPO: '/tmp/simard-engineer-loop-not-a-repo-*' — were archived as diagnostic evidence. Being rewritten each run they sort NEWEST and evict the real transcript from the bounded MAX_FILES window, manufacturing a false 'cause of death' for the stale-engineer investigator. Grounded in the archived evidence for goal advance-rysweet-agent-kgpacks-rs-to-full-parity-f29bb15c (reaped-engineers/..-1784862545), where the surfaced NOT_A_REPO came from target/operator-probe-state/engineer-loop-run/error_reflection.json while the real terminal session completed cleanly (terminal-foundation-ok, exit 0). Skip the whole target/ subtree (mirrors the existing .git skip); the genuine transcript lives outside target/ (recipe-runner logs, .claude/runtime, session state). Adds a regression test. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
📊 Coverage Summary
Coverage data from CI run. Test files matching |
This was referenced Jul 24, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Fixes the evidence-collector defect (#4449) that manufactured a false-positive stale-engineer reap of goal
advance-rysweet-agent-kgpacks-rs-to-full-parity-f29bb15c.collect_worktree_evidence(src/overseer/claim_reaper.rs) walked the entire engineer worktree skipping only.git. Cargo'starget/operator-probe-state/**andtarget/test-state/**carry cannederror_reflection.jsonfixtures. Rewritten every run, they sort NEWEST and evict the real transcript from the boundedMAX_FILES=8window — so the stale-engineer investigator saw a bogus:as the apparent cause of death, when the real terminal session had completed cleanly (
terminal-foundation-ok,COMMAND_EXIT_CODE=0,phase=complete, correct worktree cwd).Fix
Skip the whole
target/subtree during evidence collection (mirrors the existing.gitskip). The genuine engineer transcript always lives outsidetarget/(recipe-runner logs,.claude/runtime/, session state), so this is safe and cannot drop real evidence.Test
Adds
collect_worktree_evidence_skips_target_probe_fixtures: plants atarget/operator-probe-state/.../error_reflection.jsonNOT_A_REPO fixture plus a realengineer-transcript.log, asserts the fixture is excluded and the real transcript is still collected. Both evidence-collector tests pass;cargo fmt+cargo clippy --release -D warningspre-commit/pre-push gates green.Grounding
Archived evidence:
~/.simard/reaped-engineers/rysweet_Simard_advance-rysweet-agent-kgpacks-rs-to-full-parity-f29bb15c-1784862545/(manifest.json,evidence.txtline 157 +evidence-15,journal.txt). Investigation verdict: still-alive / false-positive; claim + worktree preserved (fail-closed). Related diagnostic-layer defect: #4500 (idle-age conflates completed with wedged).Closes #4449