The Agent Skills Directory

[Command Execution] (LOW): The skill's primary function is to execute shell commands (gemini, node). This is the intended behavior for interacting with local tools and is restricted to the local environment's permissions.
[Indirect Prompt Injection] (LOW): The skill is designed to read and process local files (e.g., using @path or @. syntax) which creates a surface for indirect prompt injection if those files contain malicious instructions.
Ingestion points: Local files and directories are ingested via the @ reference system in SKILL.md examples.
Boundary markers: No specific boundary markers or 'ignore' instructions are defined to separate file content from the system prompt.
Capability inventory: The skill can execute the gemini binary and node scripts, allowing for file system reads and network communication with Google's APIs.
Sanitization: No sanitization or filtering of file content is specified before the data is processed by the LLM.

gemini-cli