gws-events-subscribe

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill is designed to invoke the gws binary to subscribe to Workspace events. It executes shell commands with user-provided flags such as --target, --project, and --subscription. This is the primary intended function of the skill.
  • [PROMPT_INJECTION]: The skill facilitates the ingestion of external data from Google Workspace events (e.g., chat messages). If the agent processes this data as natural language instructions without proper isolation, it creates a surface for indirect prompt injection.
  • Ingestion points: External event stream via gws events +subscribe (SKILL.md).
  • Boundary markers: None provided in the command output handling instructions.
  • Capability inventory: Subprocess execution via the gws tool.
  • Sanitization: No sanitization logic is specified for the incoming event payloads.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 01:27 PM
Security Audit — agent-trust-hub — gws-events-subscribe