send-email
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides legitimate documentation and implementation examples for the Resend email service. All external references target official SDKs and well-known service domains.
- [INDIRECT_PROMPT_INJECTION]: The skill includes instructions for sending emails using data provided at runtime, which represents a potential injection surface for untrusted content.
- Ingestion points: Parameters for email body (html, text) and subject in
SKILL.mdand files within thereferences/directory. - Boundary markers: While the code examples use structured SDK objects, no specific prompt-level delimiters or 'ignore' instructions for the content are suggested.
- Capability inventory: Network communication with the Resend API to send emails externally.
- Sanitization: The documentation focuses on API integration and does not provide specific guidance for sanitizing email content against injection attacks.
Audit Metadata