v0.3.0 · MIT
Make every model cheaper or better. Zero-config behavioral harness for Claude Code, with a reproducible 408-run benchmark
v0.6.10 · MIT
Logic-first AI code review via semi-formal execution tracing (Premises → Trace → Divergence → Trigger → Remedy). Catches behavioral bugs, type-contract breaches & async hazards that linters miss. Six skills · Claude Code · Codex CLI · Gemini CLI.
v1.0.0 · MIT
Business consulting plugin for Claude — 16 skills, 24 commands, 5 industry overlays. Market research, competitive analysis, financial modeling, strategy frameworks, pricing, M&A, digital transformation, risk, talent, and more.
v0.3.3 · MIT
Autoharness — a self-learning skill layer for Claude Code — distills skills from your real sessions, updates them as you work, and prunes the ones that stop getting used. No daemon, no benchmark.
v1.2.1 · MIT
Eco mode for Claude Code. /eco: -31% to -73% output tokens with critical findings intact; /eco-max: up to -75% with lowered effort. Measured hardest on Claude Fable 5 (fable5), deep-studied on Sonnet 5, works on Opus 4.8 too. We publish our negative results. 82 raw benchmark runs.