The Agent Skills Directory

[SAFE]: The skill retrieves data exclusively through authorized MCP tools such as mcp__google-docs__get_doc_content, which is the standard and secure method for accessing Google Drive content.
[SAFE]: System interactions are limited to safe operations, such as creating a local storage directory (mkdir -p google-in) and writing markdown files.
[SAFE]: The skill processes content from Google Drive, which serves as an ingestion point for external data. While this creates a theoretical surface for indirect prompt injection, it is the skill's primary function. The skill mitigates risks by implementing strict filename sanitization (lowercase, alphanumeric filtering, and length limits) as defined in reference.md.
[SAFE]: The instructions regarding the AskUserQuestion bug are functional guidance designed to improve reliability on the platform and do not constitute a prompt injection or safety bypass.
[SAFE]: No obfuscated URLs, hardcoded credentials, unauthorized network operations, or persistence mechanisms were detected in the skill files.

import-gdoc