The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use standard Python development tools for formatting, linting, type checking, and testing. It specifically recommends executing commands like ruff format, ruff check, uv run mypy, and uv run pytest within the local environment. These are routine operations for software development and do not involve unauthorized privilege escalation or suspicious network activity.
[INDIRECT_PROMPT_INJECTION]: As a development aid, the skill is intended to process and interact with Python source code files which may come from untrusted external sources. While it recommends tools that execute code (such as pytest), this is a standard capability required for the skill's primary purpose.
Ingestion points: Reading and editing local *.py files and project configuration like pyproject.toml.
Boundary markers: Not explicitly defined within the skill instructions.
Capability inventory: Execution of linting and testing tools via the shell (ruff, mypy, pytest, uv).
Sanitization: None described, as the skill provides guidelines for the agent's behavior rather than input filtering.

conventions-python