The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill scans user-provided project files (styling and component files) to perform its evaluation. This creates an attack surface for indirect prompt injection, where malicious instructions hidden in code comments or design metadata could attempt to manipulate the agent's scoring logic or output format.
[COMMAND_EXECUTION]: The skill requests broad tool access (allowed-tools: ["*"]) to perform file system scanning. This is consistent with its primary purpose of analyzing local project assets but provides the agent with full access to the working directory.

design-score