Implement eval follow-ups from the #1140 run - #26
Merged
Conversation
Closes #20. Five guardrails, each pinned by a contract test citing the run's measured cost: - Decay rule: a round whose open findings are all doc/comment-only is proposed as pr-note instead of another fix round (iters 5-9 burned ~1.2M tokens polishing comments and introducing new wrong ones). Replaces the step-7 light re-review path, now deleted. - Reviewer Questions: entries route to the implementer only to verify (recorded in iter-N/question-verification.md), never as fixes to apply (~500k tokens on a built-then-reverted suggestion). - followups reads the repo's open issues and PRs so dispositions cite existing issues instead of guessing or duplicating. - Dispatch start times land in _progress.md; status unknown reports state time since dispatch (a dead implementer surfaced ~9h late). - New files enter diff.patch via git diff --no-index /dev/null <file>; git diff <ref> is silent for untracked paths. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Closes #20. Five guardrails, each pinned by a contract test citing the run's measured cost: