INSTRUCTIONS

• Look at the files in ~/.claude/learnings
• Find the highest numbered file (e.g., 0119.md)
• Increment the number by 1 and zero-pad to 4 digits (e.g., 0120.md)
• Think hard and save your learnings to that new file in ~/.claude/learnings
• Extract the title from the learning (first heading after "# " in the new file)
• Extract relevant tags from the learning content by analyzing:
  • Technology/framework names (e.g., typescript, python, react, mcp, hopper-framework)
  • Key concepts (e.g., refactoring, testing, architecture, debugging)
  • Patterns and methodologies (e.g., monoids, agent-architecture, type-safety, functional-programming)
  • Tools and libraries mentioned (e.g., bun, effection, deepeval, playwright)
  • Convert all tags to lowercase and replace spaces with hyphens
  • Include 3-7 most relevant tags
• Update ~/.learnings/README.md by:
  • Adding a new entry AFTER the "## Key Topics" section with the format: ## {number}. [{title}](learnings/{filename}) {tags}
  • The number should be the next sequential number (previous highest + 1)
  • Entries are in reverse chronological order (newest first)
  • If no specific title is found, use "Learnings" as the title
  • Tags should be formatted as backtick-wrapped hashtags (e.g., #typescript #refactoring)
  • Analyzing the Key Topics section and updating it if the new learning introduces:
    • A new major theme not yet represented
    • Significant additions to an existing topic (update the session count if applicable)
    • New frameworks, tools, or patterns that warrant inclusion
  • Keep the Key Topics section concise but comprehensive