feat(core): configurable threshold + naming taxonomy cleanup#951
Merged
feat(core): configurable threshold + naming taxonomy cleanup#951
Conversation
…configurable threshold - Rename PASS_THRESHOLD to DEFAULT_THRESHOLD with deprecated alias - Add threshold parameter to scoreToVerdict() - Thread threshold through orchestrator: evaluateCandidate → runEvaluatorsForCase → runEvaluatorList - Wire per-test execution.threshold in yaml-parser - Fix evaluate() API: add EvalConfig.threshold, pass to computeSummary() - Fix required gate to use configurable threshold instead of hardcoded 0.8 - Replace local PASS_THRESHOLD constants in benchmark-writer and artifact-writer with import - Resolution order: CLI --threshold > test execution.threshold > suite execution.threshold > DEFAULT_THRESHOLD Closes #925 (partial) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add min_score (0-1) to EvaluatorCommonSchema and all 22 evaluator config types - Parse min_score from YAML via parseRequiredAndMinScore helper - Deprecated: required: <number> now emits warning, parsed as required: true + min_score - Orchestrator required gate reads min_score preferentially over required: number - Add min_score to EvalAssertionInput (programmatic API) Closes #925 (partial) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…min_score (0-10) - Add min_score to RubricItemSchema and RubricItem type (0-1 scale) - Parser: min_score takes priority over deprecated required_min_score - Parser: required_min_score emits deprecation warning, converts to min_score - LLM grader: compare normalized score against min_score (avoids rounding issues) - Prompt generation: prefer min_score for criterion labels - Inline rubrics: support both min_score and required_min_score Closes #925 (partial) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- StudioConfig.pass_threshold → StudioConfig.threshold
- loadStudioConfig: reads studio.threshold, falls back to studio.pass_threshold, then root pass_threshold
- saveStudioConfig: writes threshold, cleans legacy pass_threshold from both root and studio
- StudioConfigResponse: add threshold field, deprecate pass_threshold
- Update all studio components to read threshold with pass_threshold fallback
- serve.ts: destructure as { threshold: pass_threshold } for minimal component diff
Closes #925 (partial)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…threshold - Regenerate eval-schema.json with min_score fields - Update rubric example: required_min_score: 8 → min_score: 0.8 - Update rubrics docs: add min_score, note required_min_score deprecation - Update skill reference: add min_score field, deprecate required_min_score - Update AGENTS.md: add live eval + agent-browser studio verification steps - Fix lint formatting issues Closes #925 (partial) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Deploying agentv with
|
| Latest commit: |
b10e335
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://34edc061.agentv.pages.dev |
| Branch Preview URL: | https://feat-925-threshold-taxonomy.agentv.pages.dev |
…Schema 2019-09 - Remove required_min_score from RubricItemSchema (Zod) and regenerated eval-schema.json - Parser still handles required_min_score at runtime with deprecation warning - Upgrade schema from JSON Schema draft-07 to 2019-09 (newest supported by zod-to-json-schema) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <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
1 participant
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.
Summary
PASS_THRESHOLD→DEFAULT_THRESHOLDwith deprecated aliasexecution.thresholdin YAML parser (resolution: CLI > test > suite > default)min_score(0-1 scale) to assertions and rubrics, deprecatingrequired: numberandrequired_min_score(0-10)pass_threshold→thresholdwith backward-compatible fallbackTest plan
--threshold 0.5correctly classifies 6/7 tests as passingstudio.threshold{"threshold":0.6}after save (notpass_threshold)studio.thresholdand cleans legacypass_thresholdE2E Evidence
Green: Live eval with configurable threshold
Threshold of 0.5 correctly used (not hardcoded 0.8). 6/7 tests pass at the lower threshold.
Green: Studio config API
Green: Config file persistence
Closes #925
🤖 Generated with Claude Code