twilio-messaging-webhooks
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [Indirect Prompt Injection Surface]: The skill processes inbound message content (
Body) which originates from external users. The documentation proactively addresses this by warning that message content is untrusted and should be isolated when passed to an LLM to prevent prompt injection. - Ingestion points: Incoming POST requests to the
/incomingendpoint extract theBodyparameter. - Boundary markers: The skill explicitly instructs developers to isolate the input and not concatenate it directly into system prompts.
- Capability inventory: The provided code handles messaging responses and configuration updates via the Twilio SDK.
- Sanitization: The skill recommends validating the
X-Twilio-Signatureheader to ensure requests originate from Twilio. - [External Dependencies]: The skill requires standard, well-known libraries (
twilio,flask,express) to function. These are common industry-standard packages for building messaging integrations and web servers.
Audit Metadata