Repository navigation
feat(evaluation): fault capture, golden set and replay runner (Phase I/II slice) - #6
Merged
Merged
Conversation
Adds evaluation_spec.md and AGENTS.md so cloud sessions can see them. .codex/ and the WIP core_graph_workspace test are intentionally left out. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Adds app/evaluation: FaultReport model + COREState adapter, JSONL golden set with signature dedupe and corrupt-line tolerance, deterministic evaluators (exact, contains, regex, max latency, forbidden substring) and a runner that replays cases through an injected callable. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_019KQ4Wi8WgwdUJKScfnBeGG
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_019KQ4Wi8WgwdUJKScfnBeGG
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What Changed
New package
backend/app/evaluation/:FaultReport(pydantic v2, JSON round-trip) plusfault_report_from_state()/capture_fault()(JSONL on disk).GoldenSet: JSONL store with stable-signature dedupe,append_fault(), and corrupt-line tolerance.run_cases()/arun_cases(): replay cases through an injected callable, return typed per-case results plus a summary.docs/evaluation/golden/seed.jsonl: 10 seed cases.docs/evaluation_spec.mdgets status and deviation notes; no roadmap box is ticked.Why
Phase I/II groundwork from
docs/evaluation_spec.md: failures become data, and a Golden Set can be replayed offline. No LLM, DB, OTEL or new services.How to Test
Notes / deviations
core/langgraph/core_graph.pyis an unused stub; the live graph iscore_graph_v2.py. A default-off hook inconversation_nodeis a follow-up.app/eval(verdict-drift harness) is untouched.Checklist
Generated by Claude Code