2‑week pilot to reduce AI‑slop PR review overhead

Contributors
1

Hi freeCodeCamp folks — I followed the recent discussion about low‑effort AI PRs and the review burden. I’m building a small tool called AI Diff Guard that flags off‑scope changes, reformat‑only diffs, and missing tests before reviewers spend time.

If you’re open to a short 2‑week pilot, I can set it up on a low‑risk repo or a subset of PRs, tune thresholds to keep noise down, and remove it if it’s not helpful.

Is there a maintainer who’d want to take a quick look, or a preferred place to try this first?

2

Hi,

Thanks for sharing this and building the tool. IMHO, we do want to preserve the human-touch to PRs, so far we think it is well under control with simple policy safeguards.

For instance we are quite certain of which PRs are noise, for instance any PR changing the README is almost always low-effort. Similarly when we see slop in other areas we simply deprioritize review and close it after a few days. The policy is simple waste our (maintainers) time, get your wasted too. And it has worked will so far.

We do not want any complexities at all to be honest. 99/100 PRs (even with AI) are actually decent.

Thanks for your offer.

Cheers, and happy coding.