Extract and categorize learnings from AI conversation sessions.
npx skills add https://github.com/mp-web3/claude-starter-kit --skill reflect-mp-web3
Please help me install this Agent Skill. Skill: reflect Source: https://github.com/mp-web3/claude-starter-kit/tree/main/skills/reflect Command: npx skills add https://github.com/mp-web3/claude-starter-kit --skill reflect-mp-web3
This Skill automatically identifies and captures valuable insights, corrections, and preferences from your AI sessions, ensuring that learned knowledge is systematically recorded and applied to improve future interactions.
Run the reflect skill to analyze the current session for learnings.
High-intent search queries and answers about installing and using this skill.
To extract learnings from AI conversation sessions, use this Skill to analyze logs and automatically identify actionable feedback, corrections, and workflow patterns. It systematically records this knowledge to improve future interactions.
Updating knowledge bases from session feedback is best handled by routing extracted insights to specific files like `rules/workflow.md` or `knowledge/user/profile.md`. The Skill intelligently directs learnings to appropriate files for persistent storage.
Yes, you can detect conflicting rules during session analysis. The Skill features conflict detection that flags contradictions between newly extracted learnings and existing knowledge or rules, prompting you to confirm necessary changes.
Yes, you need Python scripts like `extract-learnings.py` and `db.py` to categorize AI feedback and preferences. These dependencies handle data extraction and database operations required for the session analysis process.
Automated learning extraction handles workflow patterns by analyzing session interactions to identify recurring behaviors and corrections. It captures these patterns and updates your knowledge files to refine AI performance across future tasks.
A limitation of session analysis for AI improvement is that it requires Python scripts for execution and relies on predefined routing logic to specific files. It cannot dynamically create new knowledge categories outside of this established file structure.
strawgateAnalyze agent-user conversation transcripts to identify skill gaps and failure patterns.
sky770825Analyze conversational data for corrections and encode learnings into agent definitions.
caoronglinLog command failures, corrections, and workarounds into categorized Markdown files.
leifericfCapture session corrections and route them to the guidance inbox.
