How to Use the Katana MCP in Pydantic AI

Q: What happens if Katana's API changes when I'm using Pydantic AI?

Your agent will fail loudly with a ValidationError. This is a feature, not a bug. It prevents your agent from processing unexpected data structures, protecting your system from silent data corruption.

Q: How do I get started with Pydantic AI and this Katana server?

Install pydantic-ai-slimmcp, create an MCPToolset instance pointing to your Vinkius server URL, and pass it to your Pydantic AI agent. The tools are discovered and validated automatically.

Build type-safe Katana automations that never fail silently using Pydantic AI's runtime validation.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Katana MCP to Pydantic AI

Create your Vinkius account to connect Katana 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 Katana with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Create Orders with Guaranteed Correctness

When your agent calls `create_sales_order`, Pydantic AI validates the payload before sending it. If a `customer_id` is missing or in the wrong format, your code will raise a `ValidationError` immediately, instead of getting a confusing API error back from Katana. This works on the response, too. After an order is created, the data returned by Katana is parsed into a Pydantic model. If the API ever adds or changes a field unexpectedly, your agent will know instantly instead of corrupting its state.

Trust Your Inventory Numbers

Let's say you ask your agent to check stock with `get_inventory_by_variant`. You expect an integer for the quantity. Pydantic AI enforces this at runtime. If a network issue or an API change causes that tool to return a malformed string or a null value, your agent won't just crash or guess a number. It will raise a clear, specific validation error that you can catch and handle, making your inventory logic rock-solid.

A Model-Agnostic Katana MCP Server

Pydantic AI doesn't care which LLM you use. You can build your Katana agent with GPT-4, Gemini, Claude, or a local model. The tool-calling and validation logic remains exactly the same. Just add the `MCPToolset` with the Vinkius server URL to your agent's configuration. It automatically fetches the schema for tools like `list_purchase_orders` and `list_suppliers` and generates Pydantic models on the fly.

Setup guide

Set up Katana 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": {
        "katana-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

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

result = await agent.run("List recent Katana 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 Katana. 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 Katana 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 Katana MCP in Pydantic AI

It automatically validates all data sent to and received from the Katana MCP server against Pydantic models. This catches errors, like an incorrect data type for an inventory count, before they can cause problems in your application.

Yes. Pydantic AI is model-agnostic. You can use it with any major LLM provider to add a layer of type-safety to your agent's interactions with Katana tools like `list_sales_orders`.

Your agent will fail loudly with a `ValidationError`. This is a feature, not a bug. It prevents your agent from processing unexpected data structures, protecting your system from silent data corruption.

Install `pydantic-ai-slim[mcp]`, create an `MCPToolset` instance pointing to your Vinkius server URL, and pass it to your Pydantic AI agent. The tools are discovered and validated automatically.

Yes. Pydantic AI runs in your own environment, and the Vinkius MCP server provides an ephemeral, zero-trust connection. Your product, material, and customer data is processed in-memory for each request and is not stored or logged.

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