sendgrid-automation

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection vulnerability surface. 1. Ingestion points: Data entering the system through SENDGRID_IMPORT_CONTACTS (CSV processing) and SENDGRID_FILTER_ALL_MESSAGES (email activity logs). 2. Boundary markers: Absent. There are no instructions provided to the agent to treat external contact data or log content as untrusted or to ignore embedded instructions. 3. Capability inventory: The skill possesses high-impact capabilities including mass email dispatch via SENDGRID_CREATE_SINGLE_SEND, contact deletion, and sender identity modification. 4. Sanitization: Absent. No mention of data validation or escaping is included in the instructions.
  • [EXTERNAL_DOWNLOADS]: The skill requires the manual addition of an external MCP server endpoint (https://rube.app/mcp) to the agent's configuration to function.
  • [DATA_EXFILTRATION]: The skill manages highly sensitive PII, including email addresses, phone numbers, and physical addresses. It also provides tools for exporting this data externally via the SENDGRID_REQUEST_CSV and SENDGRID_DOWNLOAD_CSV tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 11:30 AM