google-workspace
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the gws binary to manage Google Workspace resources. Evidence: SKILL.md contains numerous gws commands for Drive, Gmail, and Docs.
- [PROMPT_INJECTION]: Susceptibility to indirect prompt injection exists due to the processing of untrusted external data. 1. Ingestion points: SKILL.md defines tools for reading emails (gmail users messages get) and documents (docs get). 2. Boundary markers: Absent. No instruction is provided to the agent to ignore instructions embedded in the retrieved data. 3. Capability inventory: The tool can read, write, and delete data across Google Workspace (Docs, Sheets, Gmail). 4. Sanitization: Absent. There is no evidence of data sanitization before the agent interprets the content.
Audit Metadata