skills/plurigrid/asi/julia-scientific/Gen Agent Trust Hub

julia-scientific

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill primarily serves as documentation, mapping established Python scientific libraries to their Julia counterparts without introducing malicious instructions or hidden behaviors.
  • [DATA_EXPOSURE]: Provides a functional code example for integrating with the Mathpix API to perform OCR. The script reads a local image file and transmits it via HTTP POST to https://api.mathpix.com/v3/text. This is a standard integration with a well-known scientific service and uses appropriate parameterization for credentials.
  • [REMOTE_CODE]: The skill references a large number of packages from established scientific organizations including BioJulia, FluxML, and SciML. It provides commands for installing these via the official Julia package manager (Pkg).
  • [DYNAMIC_EXECUTION]: Includes shell command examples that utilize the julia -e flag to execute package management tasks directly from the command line, which is standard practice for setting up Julia environments.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 07:04 AM
Security Audit — agent-trust-hub — julia-scientific