The Agent Skills Directory

[SAFE]: The skill provides automated configuration for a standard Python development stack using well-known tools like ruff, pytest, and pre-commit.\n- [COMMAND_EXECUTION]: Shell commands are used to initialize project structures (e.g., uv init, poetry init) and install development dependencies. These are standard operations for a project starter skill.\n- [EXTERNAL_DOWNLOADS]: The configuration templates include pre-commit hooks that fetch from established GitHub repositories, including Astral-sh (ruff) and the official pre-commit organization. These are well-known and trusted sources in the Python ecosystem.\n- [PROMPT_INJECTION]: An indirect prompt injection surface exists as the skill reads local configuration files (pyproject.toml, uv.lock) to detect the environment. 1. Ingestion points: Root directory files pyproject.toml and uv.lock. 2. Boundary markers: Absent. 3. Capability inventory: Package installation via standard managers and local file writes to config paths. 4. Sanitization: Absent. The severity is low because the skill only checks for the presence of specific sections or files to select the appropriate toolchain.

python-starter