How to Use the Unstructured MCP in LangChain

Q: How do I find out what kind of data sources Unstructured supports when using LangChain?

You use listdatasources to retrieve a manifest of all configured remote connectors. This tool shows you exactly which types of storage, like S3 or GCS, are available for your agent to process.

Q: If my workflow fails, how can I check its history with Unstructured and LangChain?

The listworkflowjobs tool lets you look at both active and historical job runs. This gives the agent a full audit trail to diagnose why the structured data transformation failed.

Q: Can I list all possible processing pipelines with Unstructured and LangChain?

Yes. The listprocessingworkflows tool provides a complete listing of every end-to-end document processing pipeline configured for the MCP Server, allowing your agent to pick the right one.

Connect Unstructured Data Pipelines: Build complex multi-step agents with LangChain.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Unstructured MCP to LangChain

Create your Vinkius account to connect Unstructured 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 Unstructured with LangChain

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Manage Sources and Destinations via MCP Server

Need to know where your data is coming from or going? Your agent can check all configured remote connectors using `list_data_sources`. It’s not just a list; it tells you if the source supports S3, GCS, and more. Similarly, before writing results anywhere, the agent runs `list_data_destinations` to confirm the target locations—whether that's a Vector DB or a specific SQL endpoint. This ensures every piece of processed data lands exactly where it should.

Observe Workflow Execution Status with LangChain

When you want your agent to run an end-to-end pipeline, you first need the blueprint. The tool `list_processing_workflows` lets the chain retrieve all available pipelines. After selecting one, checking its details via `get_workflow_details` gives visibility into the exact steps involved. For tracking, the agent monitors history by calling `list_workflow_jobs`, seeing both active and completed runs. This allows the developer to build conditional logic: if job status is 'failed', try triggering a retry using `trigger_workflow_execution`.

Query MCP Server Status for LangChain

You can't trust an agent with data unless it knows the full scope of what's available. The client uses `list_data_sources` to pull a complete manifest of all connected remote storage locations. It also checks which workflows are even possible by running `list_processing_workflows`. This allows your LangChain application to build a dynamic tool calling graph, ensuring the agent never attempts an invalid step.

Setup guide

Set up Unstructured 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 Unstructured 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({
    "unstructured-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 Unstructured 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 Unstructured. 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 Unstructured 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 Unstructured MCP in LangChain

You use `list_data_sources` to retrieve a manifest of all configured remote connectors. This tool shows you exactly which types of storage, like S3 or GCS, are available for your agent to process.

The `list_workflow_jobs` tool lets you look at both active and historical job runs. This gives the agent a full audit trail to diagnose why the structured data transformation failed.

The MCP Server exposes metadata about configured workflows and sources. You can use `get_workflow_details` to get specific configuration parameters, treating that metadata as the data you're querying.

Yes. The `list_processing_workflows` tool provides a complete listing of every end-to-end document processing pipeline configured for the MCP Server, allowing your agent to pick the right one.

This server handles the metadata and configuration details of unstructured data sources, including connection endpoints and processing workflow definitions. It doesn't touch the content itself.

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