How to Use the Cerbos MCP in Pydantic AI

Q: How do I add the Cerbos MCP server to a Pydantic AI agent?

Install pydantic-ai-slimmcp, then create an MCPToolset instance with your Vinkius server URL. Pass this toolset into the toolsets list when you initialize your Agent.

Add type-safe, validated authorization to your agent with Cerbos and Pydantic AI.

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 Pydantic AI

Create your Vinkius account to connect Cerbos to Pydantic AI 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 Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Get Guaranteed Auth Decisions

The `check_resources` tool evaluates if a principal has permission for an action on a set of resources. It's a read-only check that gives you a clear authorization result. With Pydantic AI, the response from this tool is automatically validated against a Pydantic model. If Cerbos ever returns an unexpected format, your agent will raise a `ValidationError` immediately—no silent failures or corrupted state.

Build Type-Safe Query Plans

Use `plan_resources` to have Cerbos generate a structured query plan for accessing resources. This plan tells your application exactly how to filter data for a specific user. Pydantic AI ensures the query plan your agent receives is a valid, well-formed object. You can trust the structure of the plan and use it to build database queries without worrying about malformed input from the LLM or the API.

Evaluate Policies with your Pydantic AI MCP Server

The `authzen_evaluation` and `authzen_evaluations` tools let your agent check permissions against a formal AuthZEN policy set. You can run checks one by one or batch them together. Because you're using Pydantic AI, you know the inputs you send are correctly typed and the outputs you get back are validated. This brings a level of correctness that's essential when dealing with something as critical as authorization. This MCP server handles the logic; Pydantic AI ensures the data is sound.

Setup guide

Set up Cerbos MCP in Pydantic AI

Prerequisites

Python 3.10+ installed
pydantic-ai-slim[fastmcp] package
Active Vinkius subscription with a valid endpoint token

1

Install Pydantic AI with FastMCP
Run pip install "pydantic-ai-slim[fastmcp]". The FastMCP toolset replaces the deprecated MCPServerHTTP class with full protocol support.
2

Configure the FastMCPToolset
Pass a JSON-style config dict to FastMCPToolset with your Vinkius URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. Supports Streamable HTTP, SSE, and Stdio transports.
3

Create and run your agent
Pass the toolset to Agent(toolsets=[toolset]) and call agent.run(). Swap openai:gpt-4o for any supported model — Anthropic, Google, Mistral, or Groq.

agent.py

from pydantic_ai import Agent
from pydantic_ai.toolsets.fastmcp import FastMCPToolset

toolset = FastMCPToolset({
    "mcpServers": {
        "cerbos-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

agent = Agent(
    "openai:gpt-4o",
    toolsets=[toolset],
    system_prompt="You have access to Cerbos tools.",
)

result = await agent.run("List recent Cerbos transactions")
print(result.output)

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 Pydantic AI

Install `pydantic-ai-slim[mcp]`, then create an `MCPToolset` instance with your Vinkius server URL. Pass this toolset into the `toolsets` list when you initialize your `Agent`.

Correctness. Pydantic AI's core feature is runtime type validation. When your agent gets a permission decision from Cerbos, you have a guarantee that the data structure is exactly what you expect.

Yes, automatically. The MCP toolset integration ensures that any data returned by a Cerbos tool, including the complex structure of a query plan, is parsed and validated against a Pydantic model before your agent's code ever touches it.

Yes. Pydantic AI is model-agnostic. You can use it with OpenAI, Anthropic, or a local model, and the Cerbos tool integration will work the same way.

The server processes authorization requests containing principal and resource identifiers. It does not store this data. Each request is handled in a dedicated, isolated Vinkius sandbox that is destroyed after the response is sent.

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