cloudflare-email-routing
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from incoming emails, creating a surface for injection attacks.
- Ingestion points: The skill ingests untrusted data from email headers (
message.from,message.to) and content (parsedemail.subject,email.text,email.attachments) inSKILL.mdandREADME.md. - Boundary markers: The code examples do not include boundary markers or 'ignore instructions' delimiters when interpolating email content into actions.
- Capability inventory: The skill utilizes capabilities including
message.forward(),message.reply(),env.EMAIL.send(),env.DB.prepare().run(),env.BUCKET.put(), andenv.WORKFLOW.create()across multiple patterns inSKILL.md. - Sanitization: No sanitization or validation of the incoming email body or subject is demonstrated before use in automated replies or database storage.
Audit Metadata