Summary of Unstructured API MCP Server

This repository contains an MCP (Managed Connector Platform) server implementation designed to interact with the Unstructured API, providing tools for managing sources, destinations, and workflows.

Setup:
To get started, install the necessary dependencies using uv add "mcp[cli]" and uv pip install --upgrade unstructured-client python-dotenv or uv sync. Set your Unstructured API key as an environment variable in a .env file.

Running the Server:
The server can be run using the MCP CLI with the command mcp run uns_mcp/server.py or uv run uns_mcp/server.py.

Available Tools:
The server offers a range of tools, including listing, creating, updating, and deleting sources, destinations, and workflows. Supported connectors include S3, Google Drive, Azure, Weaviate, Astra, Neo4j, MongoDB, Databricks, and OneDrive. Credentials for these connectors must be defined in the .env file.

Firecrawl Source:
The server also integrates with Firecrawl, a web crawling API, for HTML content retrieval and LLM-optimized text generation.

Intended Use Cases:
This server is intended for users who need to manage and automate data ingestion and processing workflows using the Unstructured API. It can be integrated with tools like Claude Desktop for enhanced functionality.