ln-823-pip-upgrader

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill is configured to fetch shared reference files and configurations from the author's official GitHub repository (https://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path}) if they are missing from the local environment.
  • [COMMAND_EXECUTION]: The skill executes various package management and system commands to perform its primary function, including pip install --upgrade, poetry update, pipenv update, and python -c "import <package>" for dependency verification.
  • [DATA_EXFILTRATION]: The skill reads project manifests (requirements.txt, pyproject.toml, Pipfile) and writes summary artifacts to the local filesystem (e.g., .hex-skills/runtime-artifacts/). This represents a data ingestion and writing surface that could be leveraged if the environment contains sensitive information.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes untrusted dependency files.
  • Ingestion points: Dependency manifest files including requirements.txt, pyproject.toml, and Pipfile (File: SKILL.md).
  • Boundary markers: The instructions do not specify any delimiters or safety warnings to prevent the agent from obeying instructions embedded within these external manifest files.
  • Capability inventory: The skill has the capability to install arbitrary packages, execute shell commands through package managers, and run arbitrary Python code via smoke tests (Files: SKILL.md, references/python_venv_handling.md).
  • Sanitization: There is no evidence of sanitization or validation of the input manifests before they are processed by the toolchain.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 04:39 PM