retrospective

To document lessons learned, you run a retrospective analysis that identifies outcomes, surprises, and patterns from the completed task, generating structured YAML procedure updates and observations to store in procedural memory.

A structured post-mortem reconstructs the task timeline, identifies concrete outcomes and surprises, and derives actionable procedure updates and observations to refine workflows and inform future performance.

Procedural memory capture works by generating a structured YAML payload containing the subject, date, outcome, lessons, and procedure updates from a completed interaction to codify knowledge for cross-project learning.

Yes, you can apply retrospective analysis after any completed multi-step task, conversation, or project phase to extract actionable lessons and update procedural memory for future work.

Capturing lessons learned produces a structured YAML payload containing fields for subject, date, outcome, surprises, lessons, procedure updates, and observations to be stored in procedural memory.

Retrospectives are designed for completed interactions; you should avoid using them during ongoing tasks, as the process requires a finished timeline to reconstruct events, identify surprises, and derive concrete improvements.