The Agent Skills Directory

[Remote Code Execution] (CRITICAL): The skill instructs the agent to install software by piping remote scripts directly to a shell interpreter. This allows for arbitrary code execution from a non-whitelisted source.
Evidence in SKILL.md: curl -LsSf https://astral.sh/uv/install.sh | sh and powershell -c "irm https://astral.sh/uv/install.ps1 | iex".
Evidence in references/uv-commands.md: Identical installation commands repeated.
Context: The domain astral.sh is not included in the 'Trusted GitHub Organizations' or 'Trusted GitHub Repositories' list provided in the security policy.
[Indirect Prompt Injection] (LOW): The skill is designed to process external, potentially untrusted project files which could contain malicious instructions or malformed dependency definitions.
Ingestion points: The skill reads requirements.txt, pyproject.toml, and .py files from the project directory (SKILL.md).
Boundary markers: Absent. There are no instructions to the agent to ignore or sanitize content within these files.
Capability inventory: The skill uses uv pip install, which involves network operations and file system modifications based on the content of the ingested files.
Sanitization: Absent. The agent is encouraged to execute commands directly using the names/paths found in the ingested files.

python-uv-acceleration