The Agent Skills Directory

COMMAND_EXECUTION (LOW): The skill instructions provide the agent with capabilities to execute a local script and perform file/clipboard operations.
Evidence: The skill documentation includes examples using uv run scripts/generate.py with the --output parameter and shell pipes to pbcopy.
Risk: This enables the agent to write data to any user-specified path, which could potentially result in the overwriting of sensitive system or user files if the agent does not perform path validation.

lorem-ipsum