setup
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
- [EXTERNAL_DOWNLOADS] (HIGH): The skill instructs the user to clone the
hgbrian/biomodalsrepository. This GitHub account is not within the defined trusted organizations or repositories, making it an unverified external source. - [REMOTE_CODE_EXECUTION] (HIGH): The skill facilitates the execution of scripts downloaded from the unverified repository using
modal run(e.g.,modal_boltzgen.py). Executing unverified code from external sources is a direct path to remote code execution. - [COMMAND_EXECUTION] (MEDIUM): The skill requires running multiple system commands to modify the environment, including
pip install,git clone, andmodal setup, which creates a significant attack surface if the instructions are manipulated. - [CREDENTIALS_UNSAFE] (LOW): The skill suggests running
modal token showto verify authentication. While used for debugging, this command displays sensitive authentication tokens in the terminal output, which could be captured by the agent context.
Recommendations
- AI detected serious security threats
Audit Metadata