Website and draft spec for AI Code Rating (ACR): a three-character rating, published in an ACR.md file at the root of a repo, that tells readers who maintains a project, how much of its code is written by AI, and how closely a human checks the AI's work.
| Position | Measures | Values |
|---|---|---|
| 1 | Maintainer Expertise | A (expert) – E (non-programmer) |
| 2 | AI Share | 0 (none), then 1–4 in quarters (4 = 76–100%) |
| 3 | Oversight of AI code | a (verified) – e (unchecked); a when there is no AI code |
Example: A2b means an expert maintainer, 26–50% of the code written by AI, every AI change reviewed before merge.
/— the levels, rating form, and rated projects/spec/— the latest spec; every version also lives at/spec/<version>/, listed at/spec/changelog//faq/— common questions/examples/— rating examples: common setups and the rating each one gets/related-work/— how AI Code Rating compares with other AI disclosure standards/validate/— checks anACR.mdfile in the browser
The same checks run in CI through the GitHub Action. They live in src/assets/js/acr-check.js, which the validator page uses and scripts/sync-action.mjs copies into the Action's repo.
Feedback and suggestions are welcome. Open an issue, or see CONTRIBUTING.md for how to suggest spec changes and send pull requests. Everyone taking part is expected to follow the code of conduct.
The specification is licensed under CC BY 4.0 (see LICENSE-SPEC). The site code is licensed under MIT (see LICENSE).
npm install
npm run serve # http://localhost:8080
npm run build # outputs to _website/The built site in _website/ isn't committed.
The levels live in src/_data/levels.json; the home page, spec page, and calculator are all generated from it.