AI Agent Email Inbox

Features

• Secure inbound email processing
• Webhook-based real-time notifications
• Configurable security levels (strict, domain, filtered, sandboxed)
• Webhook signature verification
• API for webhook registration

Use Cases

• Creating AI agent inboxes for automated support or task management
• Building email-to-task pipelines
• Processing untrusted inbound email programmatically
• Integrating email reception into agent workflows

Non-Goals

• Processing emails without any validation
• Trusting email headers for authentication
• Executing code directly from email content
• Mixing untrusted input directly into agent prompts without sanitization

Workflow

1. Set up Resend account and domain.
2. Choose security level for email processing.
3. Configure MX records for custom domain (if applicable).
4. Create webhook endpoint using Resend SDK or API.
5. Verify webhook signatures on incoming requests.
6. Retrieve email content using Resend SDK.
7. Apply chosen security level to validate sender and content.
8. Pass validated email content to the AI agent for processing.

Practices

• Webhook security
• Input validation
• Rate limiting
• Content filtering

Prerequisites

• A Resend account with a verified domain
• Resend API key stored in RESEND_API_KEY environment variable
• Resend SDK installed and updated for the project's language