How to Use the Drupal MCP in OpenAI Agents SDK

Q: How do I register the Drupal MCP Server with the OpenAI Agents SDK?

Install the SDK using pip, then initialize the MCPServerStreamableHttp client with your Vinkius endpoint. Pass this server instance inside the mcpservers list when instantiating your Agent. The SDK auto-discovers all 10 tools instantly.

Q: Does the OpenAI Agents SDK support caching for Drupal tool schemas?

Yes, you should set cacheToolsList=True in your Python setup code. This prevents your agent from refetching the tool definitions on every turn, reducing latency when querying listcontentnodes or listdrupalusers.

Control headless Drupal content straight from your OpenAI Agents SDK workflow with native MCP tool auto-discovery and safe guardrails.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Drupal MCP to OpenAI Agents SDK

Create your Vinkius account to connect Drupal to OpenAI Agents 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 Drupal with OpenAI Agents SDK

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Auto-discover Drupal tools in OpenAI Agents SDK

Instantly expose your headless CMS to your Python codebase. Your agents automatically map out the 10 available operations using the MCP Server HTTP transport. This means your agent can instantly run `list_content_nodes` to find active routing spaces or query specific content. You do not need to write manual wrappers or define schemas by hand. The underlying MCP Server handles the protocol translation. This lets your agent inspect active entities or fetch taxonomy structures without wasting context window tokens on custom API glue.

Safe node mutations with guardrails

Run edits on your CMS without worrying about corrupting draft states or accidentally purging live database records. The agent uses `patch_cms_node` to safely write updates to draft values while respecting strict JSON schemas. If an automated script tries to wipe a live document, your guardrails catch it before it hits the database. Destructive actions like calling `wipe_cms_node` are gated behind validation rules. This ensures that only authorized agents can drop live document entities. Your production pipelines stay clean because the MCP protocol enforces transactional integrity on every mutation.

Track taxonomy and files with tracing

Trace every taxonomy extraction and file lookup directly on your OpenAI dashboard. When your agent calls `get_taxonomy_term` to map active properties or runs `list_managed_files` to inspect storage limits, you see the exact tool execution path. This deep visibility prevents silent failures during complex content migrations. You can watch how your agent uses `get_file_metadata` to verify file storage structures in real-time. It helps you debug slow JSON:API responses and keeps your headless routing completely transparent.

Setup guide

Set up Drupal MCP in OpenAI Agents SDK

Prerequisites

Python 3.10+ installed
openai-agents package (pip install openai-agents)
Active Vinkius subscription with a valid endpoint token

1

Install the SDK

Run pip install openai-agents to install the OpenAI Agents SDK. The MCP integration is built-in — no extra dependencies needed.
2

Connect via SSE transport

Use MCPServerSse with your Vinkius endpoint URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. The SDK auto-discovers all Drupal tools at runtime.
3

Create your Agent

Pass the MCP to Agent(mcp_servers=[server]). The agent receives Drupal tools as native definitions — JSON schemas resolve automatically.
4

Run the agent

Call Runner.run(agent, prompt) to execute. The agent invokes the appropriate Drupal tools and returns structured results. Copy the full example on the right to get started.

agent.py

import asyncio
from agents import Agent, Runner
from agents.mcp import MCPServerSse

async def main():
    async with MCPServerSse(
        url="https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
    ) as server:
        agent = Agent(
            name="Drupal Agent",
            instructions="You have access to Drupal tools.",
            mcp_servers=[server],
        )
        result = await Runner.run(agent, "List recent transactions")
        print(result.final_output)

asyncio.run(main())

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 Drupal. 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 Drupal 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 Drupal MCP in OpenAI Agents SDK

Install the SDK using pip, then initialize the `MCPServerStreamableHttp` client with your Vinkius endpoint. Pass this server instance inside the `mcp_servers` list when instantiating your `Agent`. The SDK auto-discovers all 10 tools instantly.

Yes, you can configure guardrails inside your agent definition to block destructive actions. Since the server exposes `wipe_cms_node` as a standard tool, you can intercept the call in Python before it executes on your live CMS.

Yes, you should set `cacheToolsList=True` in your Python setup code. This prevents your agent from refetching the tool definitions on every turn, reducing latency when querying `list_content_nodes` or `list_drupal_users`.

You can set up specialized agents where one identifies draft nodes using `list_content_nodes` and another applies updates via `patch_cms_node`. The SDK manages the handoffs while sharing the same connection.

The server runs in an isolated Vinkius sandbox that never stores your database credentials or admin passwords. When running `list_drupal_users` or tracing `get_single_node` UUIDs, all payloads flow through encrypted, ephemeral sessions that are wiped immediately after the request completes.

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