fix: run command graders from trial cwd#56
Merged
Conversation
Full replacement of src/ and bin/ with trial runner, schemas, and CLI utilities ported from agent-build. Strips BP-specific types (no RISK_TAG, SelectionBid, TOOL_STATUS). Inlines simplified TrajectoryStepSchema with 3 step types. BREAKING: New CLI entry point, new exports structure. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
These skills taught the old pipeline's concepts (Docker evals, calibration flow, headless browser adapters) which no longer exist after the trial-runner replacement. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Gemini Code Assist MCP server config from the old pipeline. No longer relevant after the trial-runner replacement. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Old integration test wrapper for the previous pipeline's integration_tests/ directory which no longer exists. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Delete docker-compose.test.yml, Dockerfile.test, and the test-integration CI job — all referenced integration_tests/ which no longer exists. CI now runs a single test job with check + test. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
API keys were for the deleted Docker integration tests. The trial runner delegates credential management to adapters. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
AGENTS.md rewritten to reflect new trial runner architecture: updated structure diagram, skills table, capabilities, commands. Removed stale Docker test references and old skill names. Added src/tests/cli.spec.ts with coverage for: - CLI routing (trials/compare/calibrate/unknown/no-command) - parseCli meta flags (--help, -h, --schema input) - Input validation (valid JSON, invalid JSON, missing input) - Export contract verification (runTrial, schemas) All 65 tests pass. Types clean. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…/run-command-graders # Conflicts: # src/eval.ts # src/tests/eval.spec.ts
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
Tests