continuous-learning

To extract reusable patterns from Claude Code sessions, configure a Stop-hook to trigger the continuous-learning skill, which automatically evaluates and saves error resolution and debugging techniques to ~/.claude/skills/learned/.

You can capture error resolution patterns, user corrections, framework workarounds, debugging techniques, and project-specific conventions from your Claude Code sessions for rapid reuse.

Learned skills are saved locally by writing extracted patterns to ~/.claude/skills/learned/[pattern-name].md, building a categorized library of reusable markdown files across your sessions.

Yes, a Stop-hook integration is required to automate continuous learning, minimizing disruption by automatically evaluating and saving extracted patterns after each Claude Code session ends.

Yes, extracted patterns are categorized into types like error_resolution, user_corrections, debugging_techniques, and project_specific conventions, storing each as a distinct markdown file for organized reuse.

The best way to build a reusable skill library is automating extraction with a Stop-hook, which captures recurring workflows and project conventions from AI coding sessions into categorized local markdown files.