Repository navigation
feat(reaper): shadow would-close and blocked:#N verdicts, oracle claims for worktree-less sessions, and listed skips (refs theQuert/stima-api#6477) - #110
Merged
Conversation
…s, oracle claims, listed skips (refs theQuert/stima-api#6477)
… mixed-claim and row cases (refs theQuert/stima-api#6477)
… would-close verdict (refs theQuert/stima-api#6477)
… a would-close candidate Review of #110: a lowercase Done or a conditional phrase counted as a close line, and oracle_claims ran for every idle kept session before the cheap gates.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Packet 3 of theQuert/stima-api#6477 (dispatches EL-T336U / EL-T337U / EL-PATROL10: sessions falsely IDLE, falsely waiting on Leo, or DONE but never closed). Shadow only: nothing new is closed.
Survey on 2026-10-28 09:45 TPE: sec-p1 and openlux-supply had said "no open scope / the reaper can close this session" and were kept 8-10 h as
waiting, claim-unknown; auth-audit was missing from the report; no row said which issue blocks a session.oracle_claims, through the existingoracle_brief/classify_claim). They feed only the shadow verdict, sodecide()and cc-selfpush are unchanged.would-close(report only). A kept session whose only keep reasons are soft (claim-unknown, a keywordwaiting), idle past the threshold, with no dirty or unpushed work, no background child or subagent, no question, noBLOCKED:text, not PROTECT and notNEVER_CLOSE(uiux* homepage*), getswould-closewhen it has no claim and a close line, or when every claim isended.reap()still runs only with no keep reason, so APPLY never acts on it (tested with--apply).waitingonly with no claims or a claim at an oracle stop; otherwise it iswaiting-keyword, still a keep reason (cc-selfpush'sstartswith("waiting")treats it the same).blocked: #Non each kept row, fromBLOCKED:#Nin the last text or in a claim's brief.kind:*,no-tmux-field,procstart-mismatch,pane-missing,pane-ancestor-mismatch) instead of a silentcontinue.Tests:
python3 tests/session-reaper.py: 108 OK (20 new). 15 mutants killed (keyword split, person step, soft-keep gate, BLOCKED from text and brief, dirty cwd, oracle claims off, skips,anyvsallended, close line, uiux exclusion, background child, APPLY acting on would-close).Promotion of
would-closeto a close verdict comes in a later revision after SCRIPT-SAFE spot-check rounds on stima-api#6477.