ichabodcognate315
UserBuild recursive AI improvement loops to refine models, prompts, and workflows with each run
Categories
Indexed Skills (4)
evolve
Evolutionary search for agent improvement using Mind Evolution. Manages multiple parallel improvement islands via git worktrees, with selection and cross-pollination. Invoke with /evolve, "evolve my agent", "run evolutionary search", "parallel improvement".
ratchet
Autonomous ratchet loop for agent improvement. Configures optimization targets, then loops: improve agent → run agent → eval → keep or revert. Uses the /recursive-improve pipeline internally with auto-approval. Invoke with /ratchet or "run the ratchet loop", "improve my agent overnight", "autonomous improvement".
recursive-improve
End-to-end agent improvement pipeline. Analyzes raw execution traces, extracts insights, manages a skillbook, gathers domain context, defines metrics, builds a rubric, creates a prioritized action plan, presents it for review, and implements approved fixes. Trigger when the user says "improve my agent", "run the improvement pipeline", "apply insights", "/recursive-improve", or when eval/traces/ contains trace files.
benchmark
Run metric quality benchmark, store results, and compare against previous runs. Invoke with /benchmark, "run benchmark", "benchmark metrics", "check metric quality".
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.