How to Use the Cerbos MCP in Vercel AI SDK

Q: How does Cerbos work with the Vercel AI SDK?

You connect the Cerbos MCP Server to your mcpClient. Your agent can then call tools like checkresources, and the Vercel AI SDK streams the authorization results directly into your frontend components for a real-time user experience.

Q: Can I check multiple permissions at once with Cerbos and the Vercel AI SDK?

Yes. Use the authzenevaluations tool. It lets your AI client send a batch of permission checks in one request, and the results will stream back to your UI, which is ideal for complex pages.

Q: How do I check the server details?

Your agent can call the getserverinfo tool. It returns the current version and build information for the Cerbos instance you're connected to, which is useful for debugging.

Check permissions and stream access control decisions directly into your Vercel AI SDK-powered UI.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Cerbos MCP to Vercel AI SDK

Create your Vinkius account to connect Cerbos to Vercel AI SDK and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Cerbos with Vercel AI SDK

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Stream Access Verdicts to the UI

Stop showing loading spinners for permissions. Your AI client can use the `check_resources` tool to evaluate access, and with the Vercel AI SDK, the results stream directly into your React or Svelte components. Users see what they can and can't do in real time. This isn't just a boolean true/false. You get detailed outcomes for each resource. The `authzen_evaluation` tool provides a single, fast check that's perfect for verifying a specific action before a user even clicks a button. It's authorization that feels instant.

Build UIs That Adapt to User Rights

Don't fetch a giant list of items and filter it on the client. Instead, have your agent use the `plan_resources` tool. It gives you a ready-to-use query plan that only includes resources a principal is allowed to access from the start. Your UI code becomes simpler. You just run the query plan Cerbos provides and render the results. This means your application only ever requests data the user is authorized to see, which is both more secure and more efficient. This MCP Server makes that process automatic.

Check Multiple Permissions at Once

Complex interfaces often need to check many permissions to render correctly. The `authzen_evaluations` tool is built for this. Your agent can send a batch of access requests in a single call, asking Cerbos to check everything a user might want to do on a given screen. Choose the evaluation mode that fits your needs: `deny_on_first_deny` for high-security contexts or `permit_on_first_permit` for more lenient checks. The results stream back, letting your UI enable or disable features as the verdicts arrive.

Setup guide

Set up Cerbos MCP in Vercel AI SDK

Prerequisites

Node.js 18+ and a TypeScript project
ai + @modelcontextprotocol/sdk packages
Active Vinkius subscription with a valid endpoint token

1

Install dependencies

Run npm install ai @modelcontextprotocol/sdk plus your preferred model provider (e.g. @ai-sdk/openai).
2

Create the Streamable HTTP transport

Use StreamableHTTPClientTransport with your Vinkius endpoint URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com.
3

Discover and use tools

Call mcpClient.tools() to auto-discover all Cerbos tools. Pass them directly to generateText() or streamText() — no manual schema definitions needed.
4

Works with any model provider

Swap openai("gpt-4o") for any AI SDK provider — Anthropic, Google, Mistral. The MCP tools work identically across all supported models.

index.ts

import { experimental_createMCPClient as createMCPClient } from "ai";
import { StreamableHTTPClientTransport } from "@modelcontextprotocol/sdk/client/streamableHttp";
import { generateText } from "ai";
import { openai } from "@ai-sdk/openai";

const transport = new StreamableHTTPClientTransport(
  new URL("https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp")
);

const mcpClient = await createMCPClient({ transport });
const tools = await mcpClient.tools();

const { text } = await generateText({
  model: openai("gpt-4o"),
  tools,
  prompt: "List recent Cerbos transactions",
});

console.log(text);
await mcpClient.close();

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Cerbos. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Cerbos now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Cerbos MCP in Vercel AI SDK

You connect the Cerbos MCP Server to your `mcpClient`. Your agent can then call tools like `check_resources`, and the Vercel AI SDK streams the authorization results directly into your frontend components for a real-time user experience.

Yes. Use the `authzen_evaluations` tool. It lets your AI client send a batch of permission checks in one request, and the results will stream back to your UI, which is ideal for complex pages.

The main benefit is building UIs that react instantly to a user's permissions. Instead of hiding things after they load, you use `plan_resources` to only fetch what the user is allowed to see from the very beginning.

Your agent can call the `get_server_info` tool. It returns the current version and build information for the Cerbos instance you're connected to, which is useful for debugging.

The server only processes authorization data you send it for a specific check. This includes details about the principal (who is asking), the resources (what they're asking about), and the actions. This data is used ephemerally for the evaluation and is not stored.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript