Command
Use when research documents contain factual claims that need verification — law references, tax rates, costs, dates, company info, regulatory status. Use before presenting findings to stakeholders. Use after compiling data from multiple AI research runs.
Command
Use for any query involving CLAUDE.md, AGENTS.md, .cursorrules, or making a codebase work with AI coding agents. Three modes - Audit (evaluate harness maturity), Apply (add specific harness elements), Bootstrap L2 (fully automated - scan codebase, generate all missing artifacts, present for approval). Trigger on…
Command
Follow the steps in order. Don't skip any.
Command
Use after completing any task that changes code, infrastructure, services, or workflows. Works in ANY project — detects project type and runs the right post-task checklist. Keeps existing docs current. Does NOT create harness structure from scratch — use /harness-engineering for that.
Command
Track Claude Code usage by skill invocation with LiteLLM pricing.