recipe-create-meet-space

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the gws command-line utility to interact with Google Workspace services (Meet and Gmail).
  • [DATA_EXFILTRATION]: A hardcoded email address (team@company.com) is used as the recipient for outgoing emails containing meeting information.
  • [PROMPT_INJECTION]: The skill processes the output from the gws meet spaces create command and interpolates it directly into an email body. This sequence establishes an indirect prompt injection surface as there are no boundary markers to distinguish between the intended data (the meeting URI) and potentially malicious content if the command output were compromised.
  • Ingestion points: Response from gws meet spaces create (SKILL.md)
  • Boundary markers: None
  • Capability inventory: Sending emails via gws gmail +send (SKILL.md)
  • Sanitization: None
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 01:27 PM
Security Audit — agent-trust-hub — recipe-create-meet-space