sms-test-author
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes an integration pattern for processing external data via webhooks to handle SMS opt-out keywords. This creates an attack surface where untrusted input from an SMS body could influence application logic if not properly isolated.
- Ingestion points: The skill describes an endpoint
/webhooks/twilio/smsthat accepts external POST data (theBodyfield) inSKILL.md. - Boundary markers: The provided code snippets do not include markers to delimit user-supplied content from system instructions.
- Capability inventory: The documentation demonstrates performing database lookups (
User.objects.get) and state changes (updatingsms_subscribed) based on the message content. - Sanitization: No specific sanitization or validation logic is included in the simplified educational examples.
Audit Metadata