skills/cikichen/skills/python-venv/Gen Agent Trust Hub

python-venv

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Contains instructions for fetching the 'uv' package manager from its official distribution site (astral.sh) to assist with environment setup and dependency management.
  • [COMMAND_EXECUTION]: Outlines various system commands for environment lifecycle management, including environment creation, activation, and a recommendation to set the PowerShell execution policy to 'RemoteSigned' for local script compatibility.
  • [PROMPT_INJECTION]: Identifies a surface for indirect prompt injection as the skill instructs the agent to process and act upon the contents of project-specific configuration files.
  • Ingestion points: 'requirements.txt', 'pyproject.toml', 'setup.py', 'Pipfile', and 'environment.yml' located in the project directory.
  • Boundary markers: None; the skill assumes the presence of these files dictates the environment workflow.
  • Capability inventory: Employs subprocess execution for 'pip', 'uv', 'poetry', 'conda', 'python', and file system cleanup ('rm -rf').
  • Sanitization: No explicit validation or sanitization of the data within the project configuration files is described before they are passed to the underlying tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 05:35 AM