clerk-hello-world
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides code templates that ingest and display user data from the Clerk authentication service, creating a potential surface for indirect instructions to be processed by an agent interacting with the resulting application.
- Ingestion points: User identity data is retrieved via the
currentUser()andauth()functions inapp/dashboard/page.tsx,app/api/hello/route.ts, and the Express.js middleware inSKILL.md. - Boundary markers: The code examples do not include explicit instructions or delimiters to notify an agent that the rendered user data (like names or email addresses) originates from an external, potentially untrusted source.
- Capability inventory: The generated code enables reading from a remote authentication API and rendering the results in UI components or API responses, which could lead to prompt injection if the agent interprets this rendered content as instructions.
- Sanitization: The examples rely on default framework protections (such as React's JSX escaping) but do not implement specific validation or sanitization for the data retrieved from the authentication provider.
Audit Metadata