ligandmpnn

Pass

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to clone the official LigandMPNN repository from https://github.com/dauparas/LigandMPNN.git and download model weights using a provided setup script. These are legitimate resources for the tool's primary functionality.
  • [COMMAND_EXECUTION]: The workflow involves shell commands to manage dependencies (pip), retrieve external code (git), and execute inference scripts (python). These actions are transparently documented and align with the stated purpose of biomolecular design.
  • [DYNAMIC_EXECUTION]: Includes a sed command to patch a NumPy compatibility issue within the downloaded code. This is a common maintenance operation for this specific open-source package and is performed on local copies of the files.
  • [PROMPT_INJECTION]: The instructions are purely technical and instructional. There are no attempts to subvert agent safety protocols, override instructions, or hide malicious commands in the metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 1, 2026, 05:34 AM
Security Audit — agent-trust-hub — ligandmpnn