- Add step results, timing context, and build notes to LLM prompt
- LLM now sees exit codes, durations, and simple judge result
- Add guidance that long build times within timeout are acceptable
- Add separate simple/LLM verdict tracking in dual-judge mode
- Console output shows both Simple and LLM pass/fail status
- JSON summary includes separate simple/llm breakdown
- Each test report includes simplePass/llmPass fields
This helps distinguish between simple judge failures (exit code != 0)
and LLM judge failures (semantic analysis), making debugging easier.
- Add .github/workflows/build-test.yml for automated testing
- Add tests/ directory with TypeScript test runner
- Add docs/CICD.md documentation
- Remove .gitlab-ci.yml (migrated to GitHub Actions)
- Update .gitignore for test artifacts
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
