python-manager

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references installation scripts for uv and Poetry from their official, well-known domains (astral.sh and python-poetry.org). These are standard industry tools for Python development.
  • [COMMAND_EXECUTION]: Provides CLI commands for package management and environment setup. The instructions explicitly forbid direct execution, requiring the agent to ask for user consent and provide commands for manual confirmation.
  • [PRIVILEGE_ESCALATION]: Includes a PowerShell command with -ExecutionPolicy ByPass for Windows users. This is a standard installation pattern for the uv package manager and is presented as a guided reference for the user rather than an automated background action.
  • [INDIRECT_PROMPT_INJECTION]: The skill's workflow involves scanning project configuration files (pyproject.toml, requirements.txt, lock files) which are external and potentially untrusted data sources.
  • Ingestion points: Processes uv.lock, poetry.lock, pyproject.toml, and requirements.txt from the project directory.
  • Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within these files.
  • Capability inventory: The skill can generate and suggest commands for package installation and script execution based on the contents of these files.
  • Sanitization: No explicit content sanitization or validation of the project configuration files is implemented.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 04:46 AM
Security Audit — agent-trust-hub — python-manager