Analyzes merged or rejected PRs to extract concrete lessons for future contributions. Runs after PR outcomes to capture what worked (small scope, good tests, clear description) or what failed (too large, missing context, scope creep). Uses gh CLI to pull review comments, timeline, and feedback, then writes structured retrospectives to .agents/learnings with specific evidence from reviewer comments. The workflow categories feedback into style, technical, scope, testing, and process buckets, then converts observations into actionable changes. Best part is it pushes you to learn from successes too, not just rejections. Integrates with DORA metrics and treats contribution quality as a feedback loop worth measuring.
npx skills add https://github.com/boshu2/agentops --skill pr-retro