continuous-learning

Automatically extract reusable patterns from Claude Code sessions and save them as learned skills. Automatically evaluates Claude Code sessions on end to extract reusable patterns that can be saved as learned skills. This v1 skill is still supported, but continuous-learning-v2 is the preferred path for new installs. Keep v1 when you explicitly want the simpler Stop-hook extraction flow or need compatibility with older learned-skill workflows. This skill runs as a Stop hook at the end of each session: Edit config.json to customize: Add to your ~/.claude/settings.json : Homunculus v2 takes a more sophisticated approach: Key insight from homunculus: "v1 relied on skills to observe. Skills are probabilistic—they fire ~50-80% of the time. v2 uses hooks for observation (100% reliable) and instincts as the atomic unit of learned behavior." See: docs/continuous-learning-v2-spec.md for full spec. - Runs as a Stop hook at session end to evaluate transcripts, detect patterns, and save skills to ~/.claude/skills/learned/ - Configurable pattern detection across five categories: error resolution, user corrections, workarounds, debugging techniques, and project-specific conventions - Customizable via config.json with minimum session length, extraction thresholds, and pattern ignore lists - Non-blocking, lightweight approach that accesses full session context without adding per-message...