swarm-code-review

Automate multi-lane code reviews by assigning pull request surfaces to specialist agents, which then aggregate findings into a single prioritized report. This preserves evidence and deduplicates issues across lanes.

Evidence-preserving aggregation merges duplicate findings from multiple specialist review lanes while retaining source attribution. It ensures traceability and accountability by preserving the original evidence for each issue.

Coordinate specialist agents by assigning distinct review lanes for security, performance, maintainability, and style. Each lane evaluates the code independently before the aggregator merges the results.

Yes, the aggregation process merges duplicate findings identified across different review lanes. It consolidates the issues while preserving the underlying evidence and source attribution for the final report.

Generate a prioritized report by orchestrating multiple specialist agents across security and performance lanes, then aggregating their findings. The final merged output ranks issues for human decision-making.

Yes, the multi-lane code review orchestration explicitly supports maintainability and style scenarios. Specialist lanes evaluate these areas independently before merging their findings into the final report.