After a sprint or delivery, you've got bugs, user corrections, failed checks, and context blowouts scattered across session reports and feedback logs. This skill walks you through turning that mess into actual improvements: updated skill rules, new evals, workflow fixes, or tooling changes. It helps you classify each finding (skill gap, eval gap, workflow drift, etc.) and decide whether to patch an existing skill, propose a new one, or add an audit script. The four step process ends with verification, so you're not just documenting problems but closing the loop. It's structured retrospective thinking baked into a repeatable workflow instead of letting lessons learned die in a Slack thread.
npx -y skills add hoangnguyen0403/agent-skills-standard --skill retro-learn --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
supercent-io/skills-template
supercent-io/skills-template
huangjia2019/claude-code-engineering
remotion-dev/remotion
reactjs/react.dev