What problem does it solve? Non-trivial engineering work done by hand is slow, inconsistent, and impossible for a reviewer to re-verify without redoing it. This Skill enforces a working principle: build a deterministic tool (codemod, script, generator, or delegate skill) that performs or proves the work, so the artifact itself can be rerun and reviewed. ## Core Features & Use Cases - Lever-First Workflow: Do the first unit by hand to learn the recipe, then build a tool that reproduces it and diff the tool's output against the hand-done version. - Tool Selection Guidance: Choose codemods or scripts for edits, generators for repetitive files, dump-to-sqlite queries for analysis, and rerunnable checks for verification. - Subagent Contract Pattern: When fanning work out to subagents, encode the recipe, verification contract, and do-not-touch fences as a shared skill outside the delegates' write scope. - Use Case: Migrating hundreds of files to a new API? Build a codemod, prove it on one file against your manual edit, then rerun it across the codebase and commit it for future reruns. ## Quick Start Ask the AI to apply the build-the-lever principle to your next migration or bulk edit so it produces a rerunnable script instead of hand-editing files.
