clerk-webhooks
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for processing external webhook payloads, which is an ingestion point for untrusted data. However, it explicitly mitigates this risk by requiring signature verification.
- Ingestion points: Webhook payload data received at
/api/webhooksendpoints as described in SKILL.md and evals/evals.json. - Boundary markers: The instructions mandate the use of
verifyWebhookfrom@clerk/nextjs/webhooksto ensure data authenticity. - Capability inventory: Handlers typically perform database operations (Prisma) and trigger external notifications (Slack, Resend) as shown in the evaluation prompts.
- Sanitization: The skill focuses on authenticity via cryptographic signatures; standard database parameterization (e.g., via Prisma) is assumed in the generated code samples.
- [EXTERNAL_DOWNLOADS]: The skill references official documentation and development tools from Clerk, a well-known identity service provider. These references are informative and originate from trusted domains.
- Evidence: Links to
clerk.comanddashboard.clerk.comfor documentation and configuration.
Audit Metadata