Reflecting on Learnings

Extracting learnings captures patterns and techniques discovered during conversations as structured knowledge. After completing a task, invoke this Skill to analyze dialogue, identify effective workflows and tool combinations, and generate proposals for new skills or updates to existing ones that your team can reuse.

Yes. This Skill builds evaluation metadata tracking skill effectiveness over time, maintaining a history that shows which techniques and tool combinations performed best. This continuous record helps you refine workflows and identify high-impact patterns worth scaling.

Conversation analysis identifies debugging techniques, successful problem-solving sequences, effective tool interactions, and workflow optimizations that emerged during the session. These patterns become candidates for new skills or enhancements to existing ones based on proven real-world results.

Begin by invoking the Skill after task completion with a summary statement—for example, 'REFLECT on what we learned about database optimization.' The Skill then analyzes the dialogue, extracts techniques and lessons, generates evaluation metadata, and presents a structured reflection report with skill proposals.

Updating existing skills refines proven workflows with new evidence of effectiveness, while creating new skills formalizes techniques discovered for the first time. This Skill proposes both based on analysis, letting you decide which approach best serves your team's knowledge base.