The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted data from the project being refactored.
Ingestion points: The agent is instructed to read configuration files (e.g., pyproject.toml, uv.lock) and the project's source code and tests.
Boundary markers: The instructions lack explicit delimiters or warnings to ignore potential instructions that might be embedded within the code or comments of the files being analyzed.
Capability inventory: The skill performs file system operations (read/write) and executes shell commands (e.g., pytest, ruff, uv).
Sanitization: There is no evidence of sanitization or validation of the content of the external files before they are incorporated into the agent's context or used to inform command execution.
[COMMAND_EXECUTION]: The skill utilizes shell commands to verify code changes and manage project environments.
Evidence: It explicitly details the use of python -m pytest, python -m ruff, python -m mypy, and uv commands (uv run, uv sync, uv lock). These are standard professional development tools used for their intended purpose in this context.
[EXTERNAL_DOWNLOADS]: The skill uses the uv tool to manage project dependencies, which involves network activity.
Evidence: Instructions include using uv add and uv sync to manage the environment and dependencies. These operations fetch packages from official registries.

refactor-python