clerk-backend-api
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches OpenAPI specifications and repository listings from the official Clerk GitHub organization (github.com/clerk/*). These downloads are legitimate and trace directly to the vendor's own infrastructure.
- [COMMAND_EXECUTION]: Executes local shell and Node.js scripts provided within the skill to parse and extract data from the downloaded specifications. It also uses curl for API interactions and python3 for JSON formatting and parsing tasks.
- [CREDENTIALS_UNSAFE]: Uses the CLERK_SECRET_KEY for authentication. The skill instructs the agent to verify the presence of this key and sources environment variables from .env files, which is a standard and acceptable practice for secret management in developer-focused tools.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from the Clerk Backend API, such as user metadata records. Ingestion occurs during API response processing in SKILL.md. While boundary markers to distinguish data from instructions are not explicitly present, the skill mitigates risks through mandatory confirmation steps for all write and delete operations and the use of structured JSON parsing.
Audit Metadata