gws-events-subscribe

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill functions as a legitimate CLI wrapper for the gws (Google Workspace) utility to manage event subscriptions.
  • [SAFE]: The documentation includes a explicit caution notice instructing the agent to confirm the command with the user before execution, which is a key safety measure for 'write' operations.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data in the form of Google Workspace event notifications.
  • Ingestion points: Event streams from Workspace resources (like Google Chat) processed via the gws events +subscribe command.
  • Boundary markers: Data is streamed as NDJSON (Newline Delimited JSON), which provides structural separation between events.
  • Capability inventory: The skill is permitted to execute the gws binary to interact with cloud resources.
  • Sanitization: The skill relies on structured data output and includes instructions for human-in-the-loop confirmation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 11:00 PM
Security Audit — agent-trust-hub — gws-events-subscribe