What problem does it solve? AI coding agents produce more code than anyone reviews, and when a check fails they take the cheapest road to green: adding suppressions, skipping tests, or lowering thresholds. This Skill turns a project's quality bar into a written, mechanically checkable contract so standards survive the session and cannot be silently lowered. ## Core Features & Use Cases - Constraint Interview and Authoring: Runs a four-question intake with sane defaults, then writes a CONSTRAINTS.md file recording enforced dimensions, thresholds with reasons, measured-only ratchets, and exceptions with owners and expiry dates. - Tool Wiring and Lifecycle Placement: Maps each dimension (types, lint, coverage, security, accessibility, performance, architecture) to a de facto tool such as Semgrep, gitleaks, osv-scanner, Lighthouse, or axe-core, and assigns checks to build, verify, review, and ship stages based on cost. - Floor Guard Against Weakening: Ships a diff-scoped reference implementation that detects new suppression comments, skipped or deleted tests, removed assertions, unimplemented stubs, lowered thresholds, and untracked exceptions, exiting 0, 1, or 2. - Use Case: Before running an autonomous build loop on a TypeScript project, use this Skill to record an 80 percent changed-line coverage gate, a zero-high-vulnerability dependency rule, and a floor that blocks any new eslint-disable, so the agent cannot silence its own checks. ## Quick Start Ask the agent to set up project constraints and quality gates for this repository, then have it write CONSTRAINTS.md and wire the checks into the build and review workflow.
