The Agent Skills Directory

[SAFE]: The skill acts as a documentation and routing framework for Python developers, providing high-quality guidance on engineering best practices without any malicious instructions.
[EXTERNAL_DOWNLOADS]: The skill references several external resources for project setup and CI/CD. This includes the official installation script for Poetry (install.python-poetry.org) and pre-commit hooks from trusted repositories such as Astral-sh (github.com/astral-sh/ruff-pre-commit), the PyCQA (github.com/pycqa/isort), and the official pre-commit mirrors (github.com/pre-commit). These are all well-known and trusted technology providers.
[COMMAND_EXECUTION]: The modules provide instructions for using standard Python development tools including pytest, mypy, ruff, mlflow, and debugpy. These tools are executed in the context of a developer's local environment or CI/CD pipeline for legitimate quality assurance and debugging purposes.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an ingestion surface as it is designed to analyze user symptoms and queries to provide routing advice. It also includes examples of processing external data (e.g., pd.read_csv in the scientific computing module). However, it provides significant guidance on data validation, testing, and the 'Diagnosis-First' principle to mitigate risks. Ingestion points are primarily file-based data processing examples in scientific-computing-foundations.md and ml-engineering-workflows.md. Capabilities include the execution of linters, type checkers, and test suites.

using-python-engineering