ralph-deep-planner
Pass
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local scripts (
setup-ralph-deep.shandcancel-ralph-deep.sh) to manage session state and documentation. These scripts perform standard file system operations like creating directories and writing Markdown files locally within the project. - [DATA_EXFILTRATION]: No unauthorized network activity or data exfiltration attempts were detected. The skill focuses on managing local project data and state.
- [PROMPT_INJECTION]: No prompt injection, jailbreak attempts, or safety bypass instructions were found in the skill's documentation or instructions.
- [REMOTE_CODE_EXECUTION]: No remote code downloads or untrusted executions (such as
curl | bash) are present; all scripts used are part of the skill's own package.
Audit Metadata