Repository navigation
Show which role (target, auditor, judge) failed a scenario - #86
Merged
Merged
Conversation
Store a structured failure object (role required; kind, http_status, endpoint, model, retryable, attempt best effort) next to the raw error. Role comes from the engine's error_phase plus the last on_turn phase. A semantic-judge endpoint failure is now a failed scenario, not UNCERTAIN. Per-role counts in summary_metrics, run-level message, role badge, summary and filter in the UI. Results without the structure render as before. Refs #85
…stem failed The Response assessment and Judge grade cards stay for judge failures and for results without a structured failure.
There was a problem hiding this comment.
Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.
An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing.
0.6.1 records judgment.failure (role and structured detail) and caps agent-client-protocol<1 for the inspect extra. Full suite passes against the released package (1746 passed, 3 skipped).
Merged
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 #85.
Failed scenarios now record and show which role failed. Role is the only required field; the rest is best effort and works for any target or provider.
audits/failures.pybuilds a structuredfailureobject (role; kind, http_status, endpoint, model, retryable, attempt when known). The role comes from the engine'serror_phaseplus the last turn event; the target model is never assumed.infra/worker.pystores it on the result, on each failed repetition and in thescenario_failedevent, counts failures per role insummary_metrics, and the run message reads "Every scenario failed (44 target, 3 judge)".failurerender as before. The results list stacks on phones.infra/tests/test_failure_roles.py; full suite 1742 passed, 3 skipped,ruffclean on touched code.Companion engine change: SimulaMet/SimpleAudit#148, which records the role at the source. This PR works with the current engine and falls back to inference; preferring
judgment.failurewhen present is a follow-up.Not verified: no real audit against a rate-limited endpoint;
embedded_hatchettests not run. Targets that return HTTP 200 with an error as the answer text are not detected (out of scope).