skill-refinement

Evidence-based review loop for iterating AI agent skills and subagent role briefs.

Nanobot plugin for selective Hindsight memory (governed Recall/Retain) and nightly self-improving skill review via Hindsight Reflect. Fail-closed by design, single-user gated, secret-redacted.

Tunes triggers and boundaries with Tune-Skill, then polishes discovery with Findable-Polish — use it only when you want both halves at once. A triggering-only fix belongs to tune-skill, a listing-only one to findable-polish. Not for creating skills (use draft-skill).