resend-inbound
Pass
Audited by Gen Agent Trust Hub on Jul 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches email attachments from Resend's delivery infrastructure via the fetch API. Documented as part of normal workflow for handling inbound emails. Additionally suggests ngrok for local tunneling during development.
- [PROMPT_INJECTION]: The skill processes untrusted incoming email content, creating a potential surface for indirect prompt injection where malicious instructions in an email could influence agent actions. 1. Ingestion points: Webhook metadata and full email body/headers retrieved via the Resend API. 2. Boundary markers: No delimiters or instructions are used to isolate the untrusted email content in the provided code examples. 3. Capability inventory: The processing logic includes capabilities to send emails and perform network requests. 4. Sanitization: No content validation, escaping, or sanitization is implemented for the received email data.
Audit Metadata