soia-env-python-install

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill reads external project files such as pyproject.toml and requirements.txt to identify version constraints, creating a surface for indirect prompt injection.\n
  • Ingestion points: Reads pyproject.toml, requirements.txt, and .python-version from the project directory.\n
  • Boundary markers: No explicit markers or instructions are used to isolate untrusted content from these files.\n
  • Capability inventory: The skill can execute shell commands like python3, pip, brew, and pyenv, and runs a local progress recording script.\n
  • Sanitization: The skill is instructed to extract version constraints but lacks formal sanitization of the file content.\n- [COMMAND_EXECUTION]: Uses established package managers and command-line tools (brew, pyenv, pip) to perform environment setup tasks.\n- [COMMAND_EXECUTION]: Executes the local Python script scripts/record_install_progress.py to log installation status and maintain session state.\n- [EXTERNAL_DOWNLOADS]: Fetches Python binaries and package updates from well-known official sources such as Python.org and Homebrew repositories.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 08:47 AM
Security Audit — agent-trust-hub — soia-env-python-install