How to Use the Directus MCP in LangChain

Q: How does LangChain handle large payloads returned from Directus?

The agent uses listitems with limit parameters to pull manageable chunks of database records. You can configure your LangChain memory to summarize these outputs if they exceed your context window. This prevents token overflow during large database dumps.

Build multi-step reasoning pipelines where LangChain agents query and mutate your Directus database schema on the fly.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Directus MCP to LangChain

Create your Vinkius account to connect Directus to LangChain 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 Directus with LangChain

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Chain database mutations with LangChain

Your agent uses `create_item` and `update_item` via this MCP Server to write data directly to your database, feeding the outputs straight into subsequent chain links. Instead of writing custom API glue, you let the model determine when to write records based on intermediate reasoning steps. It handles the parsing and passes the payload directly to the next node in your graph. If a step fails, LangSmith tracks the exact payload passed to `delete_item` for easy debugging. This setup keeps your database operations traceable without manual logging.

Map database structures dynamically in LangChain

The model uses this MCP connection to inspect your schema using `list_collections` and `list_fields` before executing complex queries. This prevents SQL syntax errors because the agent reads the actual field configurations before trying to write or read. You don't have to hardcode schemas into your prompt templates anymore. Once the agent knows the layout, it calls `list_items` to fetch only the relevant columns. This keeps your token payload small and prevents model confusion when dealing with wide tables.

Trace user actions through LangChain chains

Your agent runs `list_activity` and `get_me` to verify who is triggering database changes during a run. This connects your LangChain context with the actual Directus user session, providing a clear audit trail for automated edits. By checking `list_roles`, the agent adapts its behavior depending on the caller's permissions. It won't attempt unauthorized writes, saving tokens and API errors.

Setup guide

Set up Directus MCP in LangChain

Prerequisites

Python 3.10+ installed
langchain-mcp-adapters + langgraph packages
Active Vinkius subscription with a valid endpoint token

1

Install dependencies
Run pip install langchain-mcp-adapters langgraph langchain-openai. The MCP adapters package converts MCP tools into native LangChain BaseTool objects.
2

Connect via HTTP transport
Use MultiServerMCPClient with "transport": "http" pointing to your Vinkius endpoint. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com.
3

Create a ReAct agent
Pass the discovered tools to create_react_agent() from LangGraph. The agent automatically routes Directus tool calls through the MCP protocol.
4

Run with any LLM
Swap ChatOpenAI for ChatAnthropic, ChatGoogleGenerativeAI, or any LangChain-compatible model. The MCP tools work identically across all providers.

agent.py

from langchain_mcp_adapters.client import MultiServerMCPClient
from langgraph.prebuilt import create_react_agent
from langchain_openai import ChatOpenAI

async with MultiServerMCPClient({
    "directus-alternative-mcp": {
        "transport": "http",
        "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp",
    }
}) as client:
    tools = client.get_tools()

    agent = create_react_agent(
        ChatOpenAI(model="gpt-4o"),
        tools,
    )
    result = await agent.ainvoke({
        "messages": "List recent Directus transactions"
    })
    print(result["messages"][-1].content)

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 Directus. 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 Directus 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 Directus MCP in LangChain

You pass your Directus static token when initializing the MCP client adapter setup. LangChain then manages the headers automatically across all tool calls. This keeps your credentials secure and separate from your agent logic.

No, this server focuses on content and metadata operations rather than structural schema modifications. Your agent uses `list_collections` to read the layout but cannot alter database tables. Use the Directus admin app to change your schema.

The agent uses `list_items` with limit parameters to pull manageable chunks of database records. You can configure your LangChain memory to summarize these outputs if they exceed your context window. This prevents token overflow during large database dumps.

The tool returns a clean error message that the agent reads and attempts to fix. If the item ID is missing, the model might try to search for the correct record using `search_items` instead. This self-healing loop keeps your chains running without crashing.

Your SQL database items and raw files remain entirely within your private Directus instance. The MCP Server acts as a local proxy, meaning your database records never pass through third-party servers. Only the specific tool outputs requested by your agent are sent to your LLM provider.

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