webhook-delivery-tester
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary purpose involves processing external webhook data, which constitutes an attack surface for indirect prompt injection. The skill provides instructions and code examples for implementing HMAC signature verification and timestamp checking to mitigate these risks.
- Ingestion points: The rawBody parameter in various handler functions within the evaluation files (e.g., src/handler.js, src/parcelo.js).
- Boundary markers: Standard Webhooks headers including webhook-signature and webhook-timestamp.
- Capability inventory: Examples simulate modifications to internal data stores like orders and credits.
- Sanitization: HMAC-SHA256 signature verification and replay window skew tolerance checks.
- [CREDENTIALS_UNSAFE]: Evaluation task files include hardcoded strings prefixed with whsec_ which represent staging or sandbox keys for testing logic. Additionally, fixture files such as test/replay/fixtures/charge-succeeded.json contain mock personally identifiable information (PII), including names, email addresses, and phone numbers, used to provide realistic test cases for sanitization logic.
- [EXTERNAL_DOWNLOADS]: The skill references external resources such as the Standard Webhooks specification and official developer documentation from vendors like Stripe, Twilio, GitHub, and GitLab for implementation guidance. These are recognized as established resources.
Audit Metadata