AI agents pass tests while producing sloppy thinking. They say "should work" without evidence. They...
AI agents pass tests while producing sloppy thinking. They say "should work" without evidence. They present partial work as complete. They embellish.
I built a tiny tool that catches this. It checks any text across four dimensions: Completeness, Consistency, Groundedness, and Honesty.
Install
pip install self-audit
Usage
To audit AI-generated tests, score how many mirror the code instead of checking it. Green CI proves...
A command-line trust layer for AI-written code: catch the cheats, prove the change meets its spec, and produce the compliance…
I run an autonomous engine that watches open-source repos for patterns we think are bugs. The...
I got tired of reviewing pull requests that looked fine until they were not. The code compiled. The...
One night, two audits, one identical score The evening of 17 May, I finish version 0.4.1...
The problem Most code an AI agent writes looks right. It compiles, it reads well, it...
