Skip to content
Get packs

Constraint-Driven Development

Writes a project's quality bar into CONSTRAINTS.md and blocks agents from quietly weakening it.

Source Download skill
Repository
addyosmani/agent-skills
Stars
112,668
Licence
MIT
Files
3 files

What it does

When installed, the agent interviews you (one question at a time, each with a sensible default) about which quality checks matter, such as test coverage, security scanning, accessibility and performance. It records the agreed bar in a CONSTRAINTS.md file at the repo root, with numeric thresholds and the command that checks each one. It also watches code changes for signs the bar is being lowered, such as new type-ignore or lint-disable comments, skipped or deleted tests, stubbed functions, or thresholds edited downward.

Great for

  • Teams using AI coding agents who want quality standards written down and enforced rather than argued about in each pull request
  • Developers starting a project or feature who have no defined quality bar yet
  • Anyone running autonomous agent loops who needs checks that the agent cannot easily silence

Skip it if

  • Throwaway scripts, spikes or prototypes, where the interview and enforced thresholds add overhead without value
  • Non-interactive runs such as CI or autonomous loops, where the skill says not to run the interview and the agent may only apply a minimal floor, leaving gaps for a human to fill
  • Projects that already have a CONSTRAINTS.md the team maintains, since the skill's guidance is to read and follow it rather than rewrite it

Try asking

  • “Set up constraints for this project”
  • “Define our standards for code quality and test coverage”
  • “Stop the agent from skipping tests or adding eslint-disable comments to get to green”
  • “Set up accessibility and performance thresholds as enforced constraints for our web app”
Step-by-step guide: how to use Constraint-Driven DevelopmentInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 3 original files, the MIT licence and an attribution note.

    constraint-driven-development.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip constraint-driven-development.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip constraint-driven-development.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Pick only the dimensions you truly care about at first; the skill suggests defaults such as test coverage and security scanning, and each pick has a cost to set up.
  • Review CONSTRAINTS.md in pull requests, since changes to the bar should be visible and deliberate.
  • Start with measured baselines rather than invented numbers, so thresholds are ones your team will actually respect.

What's inside

  • references/floor-guard.md
  • scripts/quality-gate.mjs
  • SKILL.md

Redistributed unmodified from addyosmani/agent-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.