Rung
A free audit engine for observable evidence about responsible AI coding-agent authority.
Available without a hosted service
Run Rung locally against a repository. The Python package and standalone CLI do not require Docker, a container runtime, an account, or network access to the repository being audited.
What it does
- Scores repository governance evidence across six maturity states.
- Reports a conservative maximum recommended agent authority.
- Produces deterministic JSON, text, receipt, and replay evidence.
- Remains an assessment aid, not certification, compliance advice, or release approval.
The former laptop-hosted web distribution is not required for the Rung engine and is not advertised here.