Observed arrival · 2026-10-07
AI Code Rating
An open, three-character convention for disclosing who maintains a code project, how much of its code was AI-written, and how people check that work.
- For
- Open-source maintainers disclosing AI-assisted code
- Worth noticing
- The site rates itself B4c and says Claude Code wrote nearly all its code under the author's direction.
Field notes
The page defines each position independently: maintainer expertise is based on the least experienced person who can approve a change, AI share is estimated by code lines, and oversight ranges from line-by-line verification to unchecked merges. It identifies its own rating as B4c and says Claude Code wrote nearly all of the code under the author's direction. The homepage labels the specification Draft v0.1 and links to a validator and GitHub Action.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue