ralph
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses local files (
prd.json,progress.txt,AGENTS.md) for state management and local git commands for version control, which are standard development practices. - [SAFE]: The execution loop in
ralph.shtriggers agent iterations using a local command (droid exec) with local prompt files, maintaining an internal trust boundary. - [SAFE]: The project follows a structured 'read-implement-verify-commit' workflow that includes quality checks like linting and typechecking before committing code.
Audit Metadata