What problem does it solve? AI 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 records the project's quality bar as a versioned CONSTRAINTS.md file with concrete numbers and tools, then guards it against silent erosion. ## Core Features & Use Cases - Guided constraint interview: Four questions with sane defaults capture which dimensions matter (coverage, security, performance, accessibility, architecture) and produce a working config even when the user has no numbers in mind. - CONSTRAINTS.md generation: Writes a canonical quality contract with enforced thresholds, measured-only ratchets, and tracked exceptions with owners and expiry dates. - Floor guard reference implementation: A diff-scoped Node script detects the five bar-lowering moves — new suppressions, skipped tests, removed assertions, stubs, and weakened thresholds — with exit codes 0/1/2. - Use Case: Before running an autonomous build loop on a TypeScript project, use this Skill to define an 80% changed-line coverage gate, wire gitleaks and Semgrep into CI, and ensure the agent cannot silence checks to make a failing change pass. ## Quick Start Ask the agent to set up project constraints and define our quality standards in CONSTRAINTS.md before the next autonomous build run.
