The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the ruff binary using the subprocess.run function in scripts/modern_python.py. The implementation is secure as it avoids the use of shell=True and passes arguments as a list, which prevents shell injection vulnerabilities.
[EXTERNAL_DOWNLOADS]: The skill is configured to use uvx or pipx to fetch a pinned version of the ruff tool (ruff==0.16.8) if no local version is found in the project. These are standard and trusted tools for managing Python applications.
[INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted data from the user's project files and Ruff's documentation, creating a surface for indirect prompt injection.
Ingestion points: Python source files, project configuration files (pyproject.toml, ruff.toml), and command output from the Ruff CLI.
Boundary markers: The skill does not explicitly use delimiters or specialized instructions to separate untrusted project data from the agent's core instructions.
Capability inventory: The skill has permissions to read and write files within the project and execute the ruff binary.
Sanitization: The tool uses shlex.split to safely parse environment variable overrides and Path.resolve() to normalize file paths.
[DYNAMIC_EXECUTION]: The script scripts/modern_python.py contains logic to dynamically resolve the path to the ruff executable by searching through various environments (virtualenvs, Poetry, PDM, Pixi) and system PATH. This is standard behavior for developer tools that need to integrate with local project settings.

use-modern-python