autoharness is a self-learning skill layer for Claude Code that distills reusable skills from a user's real sessions, merges near-duplicates, updates them in use, and prunes those that stop getting used — all without a daemon or an external benchmark. It fires on tool-call count rather than turns, keeps only the skills it authored, and validates a skill's worth by adherence in later turns rather than a held-out score.
- Skills are stored as plain native SKILL.md files in `.claude/skills/`; the plugin's own recall index is injected on top of the host's native mechanism
- Three distinct lifecycle signals are tracked: load (model invoked the skill), view (session read into the skill's directory), and patch (promoter landed an improvement)
- The `/learn` command allows on-demand distillation of the current session through the same proposal-and-validation chain
- Skills are stored as plain native SKILL.md files in `.claude/skills/`; the plugin's own recall index is injected on top of the host's native mechanism
- Three distinct lifecycle signals are tracked: load (model invoked the skill), view (session read into the skill's directory), and patch (promoter landed an improvement)
- The `/learn` command allows on-demand distillation of the current session through the same proposal-and-validation chain
BayramAnnakov writes about claude-reflect, a self-learning plugin for Claude Code that captures user corrections during sessions and queues them for human-reviewed application to CLAUDE.md memory files.
- Uses hooks to automatically detect correction patterns ("no, use X", "actually...", "use X not Y") and writes them to a JSON queue file
- Learnings are routed through a hierarchy of destinations: global, project-specific, personal (gitignored), modular rules with path-scoping, and low-confidence auto-memory
- Includes a `--scan-history` flag to retroactively surface missed learnings from past sessions
- Distinguishes high-confidence corrections (tool rejections, explicit "remember:" markers) from lower-confidence ones
- Uses hooks to automatically detect correction patterns ("no, use X", "actually...", "use X not Y") and writes them to a JSON queue file
- Learnings are routed through a hierarchy of destinations: global, project-specific, personal (gitignored), modular rules with path-scoping, and low-confidence auto-memory
- Includes a `--scan-history` flag to retroactively surface missed learnings from past sessions
- Distinguishes high-confidence corrections (tool rejections, explicit "remember:" markers) from lower-confidence ones
