Skip to content

scored-sh/scored

Repository files navigation

scored.sh

Every SaaS product CLI scored for AI agent readiness. One score. One grade.

What is this?

A registry that scores SaaS product CLIs (Stripe, GitHub, Resend, Vercel, etc.) on 10 dimensions of how well they work with AI agents.

Each CLI gets a score out of 100 and a letter grade (S/A/B/C/D/F).

Scoring Criteria

Criterion Max Points
JSON Output 20
Non-Interactive Mode 15
Auth Automation 15
Structured Errors 10
Agent Ecosystem 10
Schema Discovery 10
Exit Codes 5
Pagination 5
Field Selection 5
Safety Rails 5

Submit a CLI

Open an issue with the GitHub repo URL. We audit the CLI and add it to the leaderboard.

See CONTRIBUTING.md for details.

How It Works

  • Each CLI is a YAML file in data/ with booleans and enums
  • Scores are computed at build time by scripts/score.ts
  • Contributors submit data, never scores
  • The website reads computed scores from .generated/

Development

npm install
npm run dev

Related

License

MIT

About

SaaS product CLIs scored for AI agent readiness

Topics

Resources

License

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors