UBOS Asset Marketplace: Unstructured API MCP Server – Your Gateway to Streamlined Data Pipelines

In the rapidly evolving landscape of Artificial Intelligence (AI), managing and processing unstructured data is a critical challenge. UBOS understands this need and is proud to offer the Unstructured API MCP (Model Context Protocol) Server through our Asset Marketplace. This powerful tool acts as a vital bridge, connecting AI models with external data sources and tools, thereby simplifying complex data workflows and accelerating AI development.

At UBOS, we are committed to bringing the transformative potential of AI Agents to every business department. Our full-stack AI Agent Development Platform empowers you to orchestrate AI Agents, seamlessly integrate them with your enterprise data, build custom AI Agents using your preferred LLM (Large Language Model), and construct sophisticated Multi-Agent Systems. The Unstructured API MCP Server is a key component in achieving this vision, enabling efficient data handling that fuels intelligent AI applications.

Understanding the MCP Server

The MCP Server is an implementation designed to facilitate interaction with the Unstructured API. It offers a suite of tools that allows you to:

• List available sources and workflows.
• Retrieve detailed information about specific connectors and workflows.
• Create, update, and delete connectors and workflows as needed.
• Manage jobs associated with workflows.

This comprehensive functionality provides a centralized point of control for managing your unstructured data pipelines, ensuring seamless integration with your AI models.

Key Features and Benefits

• Simplified Data Integration: The MCP Server simplifies the process of connecting to various data sources and destinations, eliminating the complexities of manual integration. This allows you to focus on building and deploying AI applications rather than wrestling with data connectivity issues.
• Centralized Management: The server provides a single interface for managing all your Unstructured API interactions, including sources, destinations, workflows, and jobs. This centralized management streamlines operations and improves overall efficiency.
• Enhanced Workflow Automation: By enabling the creation, updating, and deletion of workflows, the MCP Server facilitates the automation of data processing tasks. This automation reduces manual effort, minimizes errors, and accelerates the time to insights.
• Improved Data Governance: The MCP Server provides granular control over data access and processing, ensuring that your data pipelines adhere to your organization’s security and compliance policies.
• Seamless Scalability: The server is designed to scale with your needs, supporting both small-scale experiments and large-scale production deployments. This scalability ensures that your data pipelines can handle increasing volumes of data without compromising performance.

Available Tools

The Unstructured API MCP Server comes equipped with a range of powerful tools, each designed to address specific data management needs:

• list_sources: Discovers and lists available sources from the Unstructured API, giving you a clear overview of your data landscape.
• get_source_info: Retrieves detailed information about a specific source connector, providing insights into its configuration and capabilities.
• create_source_connector: Creates a new source connector, enabling you to ingest data from a variety of sources.
• update_source_connector: Modifies an existing source connector, allowing you to adapt to changing data requirements.
• delete_source_connector: Removes a source connector, simplifying your data landscape and improving manageability.
• list_destinations: Lists available destinations from the Unstructured API, providing a clear overview of where your processed data can be sent.
• get_destination_info: Retrieves detailed information about a specific destination connector, providing insights into its configuration and capabilities.
• create_destination_connector: Creates a new destination connector, enabling you to output data to a variety of destinations.
• update_destination_connector: Modifies an existing destination connector, allowing you to adapt to changing data requirements.
• delete_destination_connector: Removes a destination connector, simplifying your data landscape and improving manageability.
• list_workflows: Lists workflows from the Unstructured API, providing a comprehensive view of your data processing pipelines.
• get_workflow_info: Retrieves detailed information about a specific workflow, providing insights into its configuration and functionality.
• create_workflow: Creates a new workflow, enabling you to define custom data processing pipelines.
• run_workflow: Executes a specific workflow, triggering the data processing pipeline.
• update_workflow: Modifies an existing workflow, allowing you to adapt to changing data processing requirements.
• delete_workflow: Removes a specific workflow, simplifying your data processing landscape.
• list_jobs: Lists jobs for a specific workflow from the Unstructured API, providing insights into the execution of your data processing pipelines.
• get_job_info: Retrieves detailed information about a specific job, providing insights into its status and performance.
• cancel_job: Cancels a specific job, allowing you to manage and control the execution of your data processing pipelines.

Connector Support

The UNS-MCP server boasts an impressive array of connectors, bridging the gap between diverse data sources and destinations. Currently, it supports:

Sources:

• S3
• Azure
• Google Drive
• OneDrive
• Salesforce
• Sharepoint

Destinations:

• S3
• Weaviate
• Pinecone
• AstraDB
• MongoDB
• Neo4j
• Databricks Volumes
• Databricks Volumes Delta Table

This extensive connector support allows you to seamlessly integrate data from a wide range of sources into your AI workflows. For a comprehensive list of supported source connectors, refer to the Unstructured platform documentation and the destination list here.

Using the Firecrawl Source

The MCP Server also supports Firecrawl, a powerful web crawling API that provides two key capabilities:

1. HTML Content Retrieval: Use invoke_firecrawl_crawlhtml to start crawl jobs and check_crawlhtml_status to monitor them. This retrieves raw HTML content from specified websites.
2. LLM-Optimized Text Generation: Use invoke_firecrawl_llmtxt to generate text optimized for large language models and check_llmtxt_status to retrieve the results. This provides clean, meaningful text content ready for AI analysis.

To utilize Firecrawl, ensure you have set the FIRECRAWL_API_KEY environment variable.

Installation and Configuration

Setting up and configuring the UNS_MCP server is straightforward:

1. Prerequisites: Ensure you have Python 3.12+ and uv installed. You also need an API key from Unstructured, which you can obtain here.
2. Installation: Install the package using uv pip install uns_mcp.
3. Configuration: Configure Claude Desktop by adding the appropriate content to your claude_desktop_config.json file (located in ~/Library/Application Support/Claude/). Refer to the documentation for specific configuration examples.
4. Environment Variables: Set your Unstructured API key as an environment variable in a .env file. Consult .env.template for configurable environment variables.

Leveraging UBOS for Enhanced AI Agent Development

By integrating the Unstructured API MCP Server into your UBOS environment, you unlock a new level of efficiency and control over your AI Agent development process. UBOS provides a comprehensive platform for:

• Orchestrating AI Agents: Seamlessly manage and coordinate the activities of multiple AI Agents.
• Connecting with Enterprise Data: Securely connect your AI Agents to your organization’s data sources.
• Building Custom AI Agents: Develop tailored AI Agents using your preferred LLM model.
• Constructing Multi-Agent Systems: Create sophisticated AI systems that leverage the power of multiple interacting Agents.

With UBOS and the Unstructured API MCP Server, you can accelerate your AI development initiatives and unlock the full potential of your data.

Why Choose the UBOS Asset Marketplace?

The UBOS Asset Marketplace is your one-stop shop for all things AI Agent development. We offer a curated selection of tools, components, and resources designed to help you build and deploy intelligent AI applications quickly and easily. When you choose the UBOS Asset Marketplace, you benefit from:

• A Wide Range of Assets: Access a diverse collection of pre-built components, models, and tools to accelerate your development process.
• High-Quality Resources: Our assets are carefully vetted and curated to ensure quality and reliability.
• Seamless Integration: Our assets are designed to seamlessly integrate with the UBOS platform.
• Expert Support: Our team of AI experts is available to provide support and guidance every step of the way.

Get Started Today

Don’t let unstructured data hold you back. Unlock the power of the Unstructured API with the UBOS Asset Marketplace’s MCP Server and transform your AI development process. Contact us today to learn more.