Your code has tests. Do your AI contributors have rules?

Repo Policy Score

4 min read Original article ↗

For maintainers reviewing AI-assisted pull requests

Your repo never told them what AI-assisted work is allowed. You find out in review.

Paste any public GitHub repository. See which contribution rules are missing, get the exact wording to fix them, and commit it. Free, no login, read-only.

No loginRead-only3 repos a day

Example scanacme/repository

AI contribution readinessLimited AI rules

Scan summary

4 contribution-policy gaps detected.

Priority gaps43 high impact

Checks passed5/2520% clear

Repo health8/98 healthy · 1 attention

Generated for youAI-POLICY.md

“Contributors must disclose in the pull request description all material AI assistance.”

Copy →

Linters read your code. Nothing reads your rules.

What you actually get

It doesn't just tell you what's missing. It writes it.

Most tools hand you a list of problems. This one hands you the files — then checks that you committed them.

01

Scan

Paste a repo URL. 25 checks against what your contribution docs actually say. Every check shows the evidence found or the public files examined.

02

Answer

A handful of questions only you can answer — where to disclose, which commands must run, who approves what. Everything else is written for you.

03

Commit

Copy or download finished Markdown, drop it in your repo, rescan. The score moves — or the tool tells you it got something wrong.

AI contribution readiness

Seven questions. One answer: can AI contribute here without guessing?

Every one of the 25 checks maps to one of them. None of it is a judgment call.

01 · Before work starts

Can work start safely?

  • Where are the rules?
  • Is AI-assisted work allowed?
02 · Before review

Can it be reviewed consistently?

  • Who approves the change?
  • What must be tested?
  • What may an agent do alone?
03 · Over time

Can governance stay current?

  • When must a human take over?
  • Who owns and updates the rules?

See how each area is scored →

Why you can check it

An AI-governance tool with no AI in it.

Deterministic

Same repo evidence, same score.

25 published rules, versioned, plain pattern matching. No model decides whether your repository “feels” ready — and when a rule was wrong, the changelog says so, and which way scores moved.

Zero access risk

No login. No write access.

No OAuth, no cloning, no code execution, no permission to change anything.

Unlisted by default

Review the result before you publish it.

Reports stay out of search. You publish when you choose.

Pro — early access

Scanning one repo once is the easy part.

Pro is being built to continuously monitor AI contribution rules and Repository Health across your organization — no pasting, no reminders. Early access members shape what it checks first.

Rescans on a schedule

Every repo in the org, without anyone starting it.

Flags policy drift

Catch missing, deleted or weakened AI contribution rules before review.

Monitors repository health

Track documentation, ownership, CI, security guidance, dependency automation and maintenance signals.

Tells you when the rules change

Know when a repository moves away from the agreed standard.

One view across the organization

See AI Readiness and Repository Health across every repo, and which ones need attention.

Shape the roadmap

What would you want it to catch?

If you look after more than a handful of repos, tell me what it should catch — early access requests decide what ships first.

Built for the people inside the review loop

If AI touches the contribution, these rules affect you.

Find the uncertainty before it becomes wasted work, inconsistent reviews or another policy argument inside a pull request.

For maintainers

Stop repeating the same AI rules in every review.

Surface missing disclosure, approval, testing and agent boundaries before the next PR — and get the wording to close them.

This is you if:
You maintain a repository and review outside or AI-assisted contributions.

For the person who owns the standard

Forty repos. Nobody checking all of them.

Free shows AI Readiness and Repository Health one repo at a time. Pro monitors both across the whole organization automatically — join early access to shape what it checks.

This is you if:
You own engineering standards across more than a handful of repositories.

Takes one paste

You will review an AI-assisted PR this week. Find out now what your repo doesn't say.

Score my repo — free →

3 repos a day · No login · Read-only