The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and implement Python code based on user-provided codebase structures and requirements, creating a surface for indirect injection.
Ingestion points: Processes codebase structures, dependencies, and implementation requirements provided by the user in the prompt context (SKILL.md).
Boundary markers: The skill instructs the agent to use specific output templates and validation steps, but lacks explicit delimiters to ignore instructions embedded within the user's code files.
Capability inventory: The skill workflow involves executing command-line tools like mypy, pytest, black, and ruff (SKILL.md).
Sanitization: No explicit sanitization or filtering of user-provided code is mentioned beyond the standard use of type-checking tools.
[COMMAND_EXECUTION]: The skill's primary workflow requires the agent to execute shell-based development tools to validate code quality and correctness.
Evidence: The 'Core Workflow' section explicitly directs the agent to run mypy --strict, black, and ruff. The testing section provides commands for running pytest (SKILL.md, references/testing.md).
[EXTERNAL_DOWNLOADS]: The skill references the author's own documentation and provides instructions for managing standard Python dependencies.
Evidence: A link to documentation on jeffallan.github.io is included in the metadata and footer (SKILL.md). This is a vendor-owned resource.
Dependencies: The packaging guide provides commands to install well-known libraries such as requests, pydantic, pytest, and mypy via pip and Poetry (references/packaging.md). These are standard industry tools.

python-pro