sendgrid-automation

Pass

Audited by Gen Agent Trust Hub on Jul 18, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill manages workflows that ingest untrusted external data, presenting an attack surface for indirect prompt injection.
  • Ingestion points: External contact data is ingested via SENDGRID_ADD_OR_UPDATE_A_CONTACT and SENDGRID_IMPORT_CONTACTS, while email content is ingested via SENDGRID_CREATE_SINGLE_SEND (all in SKILL.md).
  • Boundary markers: The instructions do not define boundary markers or delimiters to isolate untrusted data from the agent's control context.
  • Capability inventory: The skill provides capabilities for sending mass emails (SENDGRID_CREATE_SINGLE_SEND) and potentially destructive operations such as list deletion (SENDGRID_REMOVE_LIST_AND_OPTIONAL_CONTACTS).
  • Sanitization: No sanitization or validation steps for external inputs are specified.
  • [EXTERNAL_DOWNLOADS]: The skill configuration requires the connection of an external MCP server located at https://rube.app/mcp to provide its functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 18, 2026, 04:24 PM
Security Audit — agent-trust-hub — sendgrid-automation