cloudflare-email-service

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing and using official Cloudflare tools and well-known libraries.
  • References wrangler (Cloudflare's CLI) for service management and deployment.
  • Suggests the use of postal-mime and mimetext for handling email content and MIME generation.
  • Points to the official Cloudflare MCP server repository on GitHub for agent-based interactions.
  • [COMMAND_EXECUTION]: The skill includes numerous examples of command-line operations using wrangler and curl for administrative tasks such as onboarding domains, managing routing rules, and testing email delivery. These are standard development practices for the platform.
  • [DATA_EXPOSURE]: The skill correctly advises against hardcoding sensitive information, recommending the use of environment variables or Cloudflare secrets for managing API tokens (e.g., CLOUDFLARE_API_TOKEN).
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a surface area for indirect prompt injection by facilitating the processing of inbound email data by AI agents.
  • Ingestion points: The email() handler in references/routing.md and the onEmail() method in references/sending.md ingest raw email data from external senders.
  • Boundary markers: The skill suggests using the isAutoReplyEmail header to filter certain types of automated traffic, though it does not provide specific prompt delimiters for the email body.
  • Capability inventory: The associated capabilities include sending emails via the send_email binding and persisting data in Durable Object storage.
  • Sanitization: Examples demonstrate parsing MIME content using postal-mime but do not include explicit sanitization logic for the email body before it is passed to the agent context.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 09:17 AM
Security Audit — agent-trust-hub — cloudflare-email-service