remember

You can capture agent friction by explicitly describing the issue, prompting a contextual review of recent corrections, or mining session transcripts. The skill parses this input and writes structured markdown methodology notes to improve future behavior.

Session mining for agent improvement analyzes past conversation transcripts to identify recurring friction patterns. It parses the history to automatically extract behavioral corrections and codify them into structured markdown notes.

Yes, you can automatically scan recent conversations for corrections through a contextual review. The skill analyzes your conversation history, identifies friction points, and prompts you to capture them as formal methodology notes.

The best way to codify system methodology is to explicitly describe the mistake, such as issuing a command to remember a specific constraint. The skill then generates a structured markdown note to prevent the agent from repeating the error.

Yes, session mining requires past session transcripts as input to identify recurring friction patterns. The skill parses these transcripts to extract behavioral corrections and write them into structured methodology notes.