We scanned 2,204 merged AI-agent PRs. Zero declared what they intended to change.

I spent a couple of weeks scanning merged pull requests opened by coding agents — Devin, GitHub Copilot's coding agent, Codex, Claude Code, and Cursor — across public repositories. 2,204 of them, read-only through the GitHub API, no LLM judgment anywhere, just deterministic checks.

The number that stuck with me: zero of the 2,204 declared, in any machine-checkable form, what they intended to change. Not "most didn't" — zero.

Of the 349 that touched CI workflows or package manifests, 12.9% escalated workflow permissions and 17.5% added unpinned actions. 3.9% edited the files that steer every future agent run (AGENTS.md, CLAUDE.md, .mcp.json).

Most of those are probably fine. That is sort of the point: nobody declared them, so nobody can tell the difference without reading every diff by hand.

So I'm genuinely curious before launch day — how is your team handling agent-opened PRs right now? The same review path as a human PR? A separate label and a slower lane? Or has nobody really looked yet?

Methodology and every query are public if anyone wants to poke holes in it: https://github.com/sjh9714/mergewarden/blob/main/docs/study/methodology.md — happy to be told the sample is wrong.