gworkspace
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes content from external Google Workspace sources (such as emails, documents, and spreadsheets) which may contain hidden or malicious instructions designed to influence the agent's behavior. \n
- Ingestion points: Tools auto-discovered from the
workspace-mcpserver and methods likeget_doc_contentdocumented inSKILL.md. \n - Boundary markers: None present in the provided skill code or documentation to delimit external data from system instructions. \n
- Capability inventory: Provides read and write access to Gmail, Google Drive, Docs, Calendar, Sheets, and Slides. \n
- Sanitization: The integration layer in
src/gworkspace/__init__.pydoes not implement any validation, filtering, or escaping of the content returned by the MCP server.
Audit Metadata