Exclude agent reasoning from findings judge input - #54
Merged
Merged
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 findings judge currently sees agent reasoning summaries as trajectory evidence and can penalize speculative drafts that were never delivered. Exclude top-level
thinking:andreasoning:steps when constructing judge messages.Original step numbers remain attached to tools and screenshots. Saved traces, final answers, output files, scoring rules and deterministic canary checks stay intact. Filter before the trajectory text budget; record adapter version 2.1.4. Existing results and charts are not recomputed.
Validation: 72 offline tests completed (69 passed, 3 optional integration tests skipped), including screenshot-reference and raw-evidence preservation; rubric revision review and the workflow's Ruff checks passed. No paid model or browser evaluations.
Summary by cubic
Prevents the findings judge from treating top-level
thinking:andreasoning:summaries as trajectory evidence, so speculative drafts no longer affect scoring. Saved traces, tool and screenshot evidence, final answers, output files, and deterministic checks remain unchanged.2.1.4.Written for commit fec2232. Summary will update on new commits.