caveman
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill uses instructions to maintain a persistent persona across multiple interaction turns once triggered. It includes clear exit conditions for the user to disable the mode ("stop caveman" or "normal mode"), which mitigates risks of unintended instruction persistence.
- [COMMAND_EXECUTION]: The skill provides three Python utility scripts (compressor, linter, and estimator) intended for text processing. These scripts are strictly limited to the Python standard library and do not perform network operations, access sensitive files, or execute external commands.
- [EXTERNAL_DOWNLOADS]: Reference documentation links to an official GitHub repository by Matt Pocock, a well-known developer. These references are used for attribution and technical guidelines and do not involve the execution of untrusted remote code.
Audit Metadata