What problem does it solve? AI coding agents produce more code than anyone reviews, and without a written quality bar they quietly lower standards — adding suppressions, skipping tests, or weakening thresholds to reach green. This Skill turns a project's quality bar into a written, mechanically checkable contract that outlives the conversation. ## Core Features & Use Cases - Constraint Interview & Defaults: Runs a four-question intake with sane defaults, then writes a CONSTRAINTS.md file with numbered thresholds, measured ratchets, and tracked exceptions. - Tool-Backed Enforcement: Maps each dimension (types, lint, coverage, security, performance, accessibility, architecture) to a de facto tool like tsc, Semgrep, gitleaks, Lighthouse, or axe-core, with lifecycle placement from edit loop to CI. - Floor Guard Reference: Ships a diff-scoped reference implementation that detects weakened thresholds, skipped tests, new suppression comments, unimplemented stubs, and surprise exceptions. - Use Case: Before running an autonomous build loop on a TypeScript project, use this Skill to record an 80% changed-line coverage gate, a zero-high-vulnerability dependency rule, and a floor banning new eslint-disable comments, then wire them into check:fast and check:full scripts. ## Quick Start Ask the agent to set up project constraints and define our quality standards in a CONSTRAINTS.md file with enforced coverage and security checks.
