ai-gateway
Pass
Audited by Gen Agent Trust Hub on Apr 14, 2026
Risk Level: SAFE
Full Analysis
- [External References and Downloads]: The skill references documentation, sitemaps, and repositories located on Vercel's official domains (vercel.com, ai-sdk.dev) and GitHub. These are used to provide the agent with authoritative and up-to-date information regarding model routing and provider configuration.
- [Package Management]: The instructions include standard commands for installing official Node.js packages, such as
@ai-sdk/gateway, using common package managers like npm, pnpm, bun, and yarn. This is consistent with the skill's purpose as a developer tool. - [Credential Management Configuration]: The skill provides guidance on setting up authentication using environment variables (
AI_GATEWAY_API_KEY) and specifically promotes the use of Vercel's OIDC-based authentication (VERCEL_OIDC_TOKEN) via thevercel env pullcommand. This approach aligns with modern security best practices for secret management by favoring short-lived tokens over static keys. - [Standard CLI Operations]: The skill includes usage examples for the Vercel CLI to link projects and synchronize environment variables. These are routine operations for developers working within the vendor's ecosystem.
- [Data Ingestion Surface]: As an AI gateway integration, the skill describes patterns for passing user-provided content to external AI models. While this involves handling untrusted data, the skill includes patterns for handling error responses and managing usage limits, which are typical for this architectural role.
Audit Metadata