Their methodology section says they scan for authors and commit messages indicating AI use. This wouldn't work on any of my repos, for example, because I review and commit everything manually.
Overall, I find this sort of thing distasteful, but to each their own.
Here's what they say:
In the tools file, there's a big list of different LLM agents and tooling. Repo-slopscore analyzes every commit in the history, looking at
Committer name
Committer email address
Commit messages
in various ways. If AI activity signals are detected here, they'll get pointed out. The tools file covers a good amount of known (to me, at least) AI agents, and is made even more accurate by broad-spectrum filtering for phrases such as "generated with/by/using", which are telltale signs of LLM activity.
Fairly brittle approach: https://codeberg.org/polyphony/repo-slopscore
Their methodology section says they scan for authors and commit messages indicating AI use. This wouldn't work on any of my repos, for example, because I review and commit everything manually.
Overall, I find this sort of thing distasteful, but to each their own.
Here's what they say:
In the tools file, there's a big list of different LLM agents and tooling. Repo-slopscore analyzes every commit in the history, looking at
in various ways. If AI activity signals are detected here, they'll get pointed out. The tools file covers a good amount of known (to me, at least) AI agents, and is made even more accurate by broad-spectrum filtering for phrases such as "generated with/by/using", which are telltale signs of LLM activity.So they care more about the slopee than the slop.