The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides installation instructions for the uv CLI using a script download and execution pattern.
Evidence: SKILL.md contains the command curl -LsSf https://astral.sh/uv/install.sh | sh. This is a well-known installation method for the Astral uv tool.
[EXTERNAL_DOWNLOADS]: The skill facilitates downloading and installing Python packages from external registries.
Evidence: Multiple files (e.g., SKILL.md, references/dependency-management.md) demonstrate the use of uv pip install to fetch packages from PyPI and external Git repositories.
[COMMAND_EXECUTION]: The skill relies on the Bash tool to execute environment management and package installation commands.
Evidence: The allowed-tools field in SKILL.md specifies Bash, which is used throughout the documentation to demonstrate uv commands.
[INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface where the agent processes external files that could contain malicious instructions.
Ingestion points: The agent reads pyproject.toml, requirements.txt, and requirements.in files as part of its dependency management tasks.
Boundary markers: No explicit boundary markers or instruction-ignoring warnings are present in the documentation for handling these files.
Capability inventory: The skill utilizes Bash to run uv pip install, uv run, and uv sync, providing a pathway for code execution based on processed file content.
Sanitization: There is no evidence of sanitization or validation of the contents of requirement or configuration files before they are passed to the uv tool.

python-env