kaizen

To capture lessons from a sprint retrospective, you identify what went well, what didn't, and recurring patterns during development iterations, then record these actionable insights into internal AI state files for continuous improvement.

A V-phase retrospective is a structured review for development sprints that surfaces successes, failures, and recurring patterns. You need it during product iterations to capture actionable lessons and drive continuous process improvement.

You update .ai_state/lessons.md by running a V-phase retrospective that identifies successes, failures, and recurring patterns from development sprints. The process automatically records these actionable lessons and new insights into the file.

Yes, you can track recurring agent errors by running a V-phase retrospective that identifies recurring patterns and failures from development iterations. The process specifically implements targeted updates to the Agent Errors section in .ai_state/conventions.md.

Yes, the retrospective process generates a concise summary for stakeholders after capturing actionable lessons and updating AI state files. This summary informs future development sprints and product iterations based on the reviewed V-phase outcomes.

The best way to drive continuous process improvement is running structured V-phase retrospectives that capture successes, failures, and recurring patterns from development sprints. This updates internal AI state guides to prevent past mistakes and refine development conventions.