One contributor opened 106 PRs in a day, 3 seconds apart

When a repo gets 3,400 pull requests a week

Greptile's post treats OpenClaw, the fastest-growing repo in GitHub history, as a natural experiment in AI-authored contributions. The volume went from 2 PRs a week in December to 3,400 by February. The merge rate fell from roughly 48% to 9.3%. One contributor filed 106 PRs in a single day, with a median of three seconds between submissions. Four separate people opened the identical PR with the same SearXNG title.

The pattern in the data is the headline argument: "the thinking matters a lot more than the typing." Contributions that require deep understanding of the existing codebase outperform novel features by nearly 4x (refactors merge at 35%, features at 9%). Repeat contributors do better too: 8.2% for first-timers, 18.6% once someone has five or more PRs. The proposed fix is borrowed from the inbox wars of the early 2000s: reputation infrastructure and trust tooling, citing Mitchell Hashimoto's Vouch.

The thread doubts the analogy it's built on

The sharpest pushback came on the spam metaphor itself. An engineer pointed out that email defenses never tracked individuals at all.

"with email it was always about sender reputation based on email servers (IPs)… But never about individual users. It's the organizations running the email server, who make sure users behave." — an engineer on how email reputation actually worked

Others found the whole arms race absurd. If you can build an agent to screen the spam, why not skip a step.

"Wait. So to combat AI spam there's AI agents to prevent it? Why can the anti spam agents not just do the work directly???" — a commenter on the recursion problem

And the irony was not lost on the crowd: the complaint is coming from inside the house.

"the irony isn't lost that this article is specifically complaining about AI slop PRs to the Open Claw repo. If the maintainers are that tired of it, they should update OpenClaw to prevent it from submitting PRs to their repo." — a commenter on the self-inflicted flood

A dissenter rejected the spam frame entirely, arguing the flood is mostly new programmers worth keeping.

"Vast majority of contributors made a change useful for themselves that they wish to share… It's better to think of this as an influx of new programmers." — a commenter reframing the flood as onboarding

The unresolved part is enforcement. Email reputation worked because a few large operators policed millions of users to keep their IPs clean. GitHub has no such intermediary, which is why every proposed fix ends up asking maintainers to do the policing the protocol can't.