pymol

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands using uv run to launch Python scripts for molecular rendering.
  • [DYNAMIC_EXECUTION]: The skill operates by generating Python scripts from templates and executing them at runtime to perform specific visualization tasks.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external molecular structure files (e.g., .pdb, .cif) which represent an untrusted data ingestion surface.
  • Ingestion points: Structure files are loaded using the cmd.load() command within the generated Python scripts.
  • Boundary markers: No explicit boundary markers or 'ignore' instructions are provided when loading these external files.
  • Capability inventory: The skill utilizes subprocess execution via uv run and has the capability to read and write files on the local filesystem.
  • Sanitization: There is no explicit sanitization or validation of the structure file content prior to processing by the PyMOL engine.
  • [EXTERNAL_DOWNLOADS]: The skill downloads and installs the pymol-open-source-whl package via uv to provide the required molecular modeling libraries.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 02:08 PM
Security Audit — agent-trust-hub — pymol