Autoresearch/scoring mar27#4
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
|
Caution Review failedThe pull request is closed. ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: ⛔ Files ignored due to path filters (2)
📒 Files selected for processing (22)
📝 WalkthroughWalkthroughAdded comprehensive autoresearch framework including fixture-based evaluation for scoring and routing tracks, manifest configuration, evaluator scripts, a git-integrated runner for automated iteration loops, performance benchmarking with Puppeteer, and supporting documentation and tests. Changes
Sequence Diagram(s)sequenceDiagram
participant Runner as Autoresearch Runner
participant Git as Git Operations
participant Evaluator as Track Evaluator
participant Logger as Result Logger
participant TSV as results.tsv
Runner->>Runner: Parse --track argument
Runner->>Git: Fetch branch name & changed files
Git-->>Runner: Branch, changed files
Runner->>Runner: Validate files in track allowlist
Runner->>Git: Stage changed files
Runner->>Git: Commit with track-prefixed message
Git-->>Runner: Commit hash
Runner->>Logger: Read previous best metric
Logger->>TSV: Query best "keep" metric
TSV-->>Logger: Best metric (if exists)
Logger-->>Runner: Previous best
Runner->>Evaluator: Evaluate fixture set
Evaluator-->>Runner: Report with metric & failures
Runner->>Runner: Compare metric vs previous best
alt Metric improved or no previous best
Runner->>Logger: Append "keep" result row
Logger->>TSV: Write result
Runner->>Logger: Write JSON report
else Metric did not improve
Runner->>Git: Reset HEAD to parent
Runner->>Logger: Append "discard" result row
Logger->>TSV: Write result
Runner->>Logger: Write JSON report
end
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Poem
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Summary by CodeRabbit
Release Notes
New Features
scoringandroutingtracks with constrained file edits and automated evaluation.Documentation
Chores