google-workspace
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the
@googleworkspace/clipackage from the official NPM registry as a global dependency. This utility is a well-known tool for managing Google Workspace environments via the command line. - [COMMAND_EXECUTION]: The skill facilitates the execution of various
gwscommands to interact with Google Drive, Gmail, Calendar, and other services. It correctly guides the user to use--dry-runflags for testing and provides clear instructions for pipeline processing usingjq. - [PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection because it allows an AI agent to read content from external sources (Gmail messages, Drive files) and provides tools to execute actions based on that content.
- Ingestion points: Untrusted data enters the agent's context through methods like
gws gmail users-messages get,gws drive files get, andgws sheets spreadsheets-values getas documented inSKILL.md. - Boundary markers: The provided instructions do not include specific delimiters or 'ignore' instructions to prevent the agent from mistaking data content for instructions.
- Capability inventory: The skill enables high-privilege operations including sending emails (
gmail users-messages send), deleting files (drive files delete), and modifying permissions, which could be exploited if an injection occurs. - Sanitization: There is no mention of sanitizing or escaping the retrieved data before it is processed by the agent or passed to subsequent commands.
Audit Metadata