The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized behaviors were identified. The skill instructions and references align with the legitimate purpose of providing developer documentation for a trusted tool.
[INDIRECT_PROMPT_INJECTION]: The skill describes functionality for a tool that processes external content, which presents a potential injection surface.
Ingestion points: Documentation in references/documentation_index.md mentions web_fetch for web pages and multi-file tools for reading local project files.
Boundary markers: No specific boundary markers or 'ignore' instructions for external data were found in the provided files.
Capability inventory: The referenced tool supports shell command execution (run_shell_command), file system modification, and memory persistence (save_memory).
Sanitization: No specific sanitization or validation logic for external content is described in the skill.

gemini-cli