expose-agentkit-mcp

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external data (emails via gmail_fetch_mails) and uses this data to trigger subsequent actions (creating calendar events via googlecalendar_create_event). This creates a surface for indirect prompt injection where malicious instructions embedded in an email could attempt to influence the agent's behavior. * Ingestion points: External email content retrieved in Step 3/5. * Capability inventory: File-write/action capability via calendar event creation. * Boundary markers: Not specified in the provided implementation examples. * Sanitization: The skill relies on the underlying LLM's default processing rather than explicit sanitization of the fetched email body.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 08:09 AM
Security Audit — agent-trust-hub — expose-agentkit-mcp