python-uv-workflow

Pass

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands using the uv toolchain (e.g., uv run, uv add, uv sync) and pre-commit (e.g., pre-commit install). These are standard development operations for managing Python environments.
  • [EXTERNAL_DOWNLOADS]: The .pre-commit-config.yaml template references external repositories (github.com/pre-commit/pre-commit-hooks and github.com/jendrikseipp/vulture) to fetch hook configurations. These are widely used and reputable resources in the Python community.
  • [PROMPT_INJECTION]: The SKILL.md file contains a "Read Acknowledgement" instruction requiring the agent to output a specific phrase: "ፀ 祡は python-uv-workflow を読みました。". This is a behavioral instruction and does not attempt to bypass safety guidelines or override system prompts.
  • [SAFE]: The skill is designed to process and refactor existing Python projects, which presents a surface for indirect prompt injection. However, since it utilizes standard development tools and operates within a typical development workflow, this is assessed as a standard operational risk.
  • Ingestion points: Reads project files such as Makefile, .pre-commit-config.yaml, and Python source code.
  • Boundary markers: Not explicitly defined in the provided instructions.
  • Capability inventory: Can execute shell commands via uv, install packages, and write/modify configuration files.
  • Sanitization: Relies on the default behavior of the uv and pre-commit tools when handling file content.
Audit Metadata
Risk Level
SAFE
Analyzed
May 6, 2026, 08:39 PM
Security Audit — agent-trust-hub — python-uv-workflow