Skills
skills/1mangesh1/dev-skills-collection/python-env/Gen Agent Trust Hub

python-env

Fail

Audited by Gen Agent Trust Hub on Feb 15, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • REMOTE_CODE_EXECUTION (CRITICAL): The skill promotes the dangerous 'curl | bash' anti-pattern for installing Poetry, uv, and pyenv. Evidence in SKILL.md: https://install.python-poetry.org piped to python3, https://astral.sh/uv/install.sh piped to sh, and https://pyenv.run piped to bash. These domains are not in the Trusted External Sources whitelist.
  • COMMAND_EXECUTION (HIGH): Shell scripts and Python utilities execute system commands using variables derived from user input without sanitization. Evidence: scripts/venv-manager.sh interpolates $env_name directly into shell commands, and scripts/poetry-manager.py uses unvalidated package names in subprocess.run calls.
  • PROMPT_INJECTION (HIGH): Vulnerable to Indirect Prompt Injection (Category 8). The skill ingests external content (package and environment names) and possesses high-privilege execute and write capabilities without boundary markers or sanitization logic (Evidence in scripts/venv-manager.sh and scripts/poetry-manager.py).
  • EXTERNAL_DOWNLOADS (MEDIUM): Systematic reliance on external package registries (PyPI, Conda) for dependency management throughout SKILL.md and referenced guides without explicit version pinning or integrity verification in all examples.
Recommendations
  • CRITICAL: Downloads and executes remote code from untrusted source(s): https://pyenv.run, https://install.python-poetry.org, https://astral.sh/uv/install.sh - DO NOT USE
  • AI detected serious security threats
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 15, 2026, 02:05 AM