Observed arrival · 2026-08-26
Agent Readme grades repositories for AI-agent readiness
A public tool and local CLI that score GitHub repositories on instructions, setup, tests, context economy, and navigability.
Field notes
The assessment separates repository readiness into five weighted categories, with instructions and a verification loop accounting for more than half the score. Its sample covers 867 highly starred software repositories, where the homepage reports a median of 62 and finds that 38% lack a test command an agent can discover. Public reports can be requested as plain text, while the local CLI is presented for private repositories and supports drafting an AGENTS.md file or failing a build below a selected minimum score.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue