UBOS MCP Server: Bridging the Gap Between LLMs and the Web

In the rapidly evolving landscape of AI, Large Language Models (LLMs) are proving to be invaluable tools for a multitude of applications. However, their effectiveness is often limited by their ability to access and interact with real-time information and external resources. This is where the Model Context Protocol (MCP) comes into play, and UBOS’s MCP Server stands out as a powerful solution.

The UBOS MCP Server acts as a vital bridge, enabling LLMs to seamlessly connect with and utilize web-based tools. By providing a standardized interface, it allows LLMs to leverage the vast amount of information available on the internet, enhancing their capabilities and expanding their potential use cases.

What is MCP and Why is it Important?

The Model Context Protocol (MCP) is an open standard designed to streamline how applications provide contextual information to LLMs. Think of it as a universal language that allows LLMs to understand and interact with different tools and data sources. Without MCP, integrating LLMs with external resources can be a complex and time-consuming process, requiring custom code and integrations for each tool.

MCP solves this problem by providing a standardized way for applications to expose their functionality to LLMs. This allows developers to easily create tools that LLMs can use, and it allows LLMs to access a wider range of information and capabilities.

Key Features of the UBOS MCP Server

The UBOS MCP Server offers a range of features designed to make it easy to integrate web-based tools with LLMs:

• Web Scraping Tool: The core functionality currently provided is the web_scrape tool. This tool fetches the content of a specified URL and returns it in markdown format, ready for LLMs to process. It prefixes the URL to use r.jina.ai for content retrieval.
• Easy Installation: Setting up the UBOS MCP Server is straightforward. With simple git clone and pip install commands, you can quickly get the server up and running.
• Flexible Deployment: The server can be run using standard input/output (stdio) or Server-Sent Events (SSE) transport, offering flexibility in deployment options.
• MCP Inspector Compatibility: The server is fully compatible with the MCP Inspector, a tool for testing and debugging MCP servers. This allows you to easily test and verify the functionality of your tools.
• Claude Integration: The server is designed to seamlessly integrate with Claude for Desktop, allowing you to use web-based tools directly within the Claude interface.
• Extensible Architecture: Adding new tools to the server is a simple process. You can easily create new Python files in the tools/ directory and register them with the MCP server.
• Streamlit UI: A Streamlit-based UI is included, allowing you to connect to and test all your MCP servers configured in Claude for Desktop. This provides a visual interface for interacting with your tools and verifying their functionality.

Use Cases for the UBOS MCP Server

The UBOS MCP Server unlocks a wide range of use cases for LLMs, including:

• Enhanced Information Retrieval: LLMs can use the web_scrape tool to fetch information from websites and incorporate it into their responses. This allows them to provide more accurate and up-to-date information.
• Content Summarization: LLMs can use the web_scrape tool to fetch the content of a website and then summarize it for the user. This can be useful for quickly understanding the main points of a long article or web page.
• Research and Analysis: LLMs can use the web_scrape tool to gather data from multiple websites and then analyze it to identify trends and patterns.
• Web Automation: LLMs can use the UBOS MCP Server to interact with web applications and automate tasks such as filling out forms, submitting data, and navigating websites.
• Real-time Data Integration: Integrate live web data into AI workflows for dynamic analysis and decision-making.
• Contextual Customer Support: Equip AI agents with the ability to fetch real-time information from a company’s website or knowledge base to provide accurate and helpful customer support.

Integrating with Claude for Desktop: A Step-by-Step Guide

One of the most compelling features of the UBOS MCP Server is its seamless integration with Claude for Desktop. Here’s how you can set it up:

1. Install Claude for Desktop: Ensure you have Claude for Desktop installed on your system.

2. Locate the Configuration File: Find the claude_desktop_config.json file. The location varies depending on your operating system:

   • Mac: ~/Library/Application Support/Claude/claude_desktop_config.json
   • Windows: %APPDATA%Claudeclaude_desktop_config.json

3. Add the MCP Server Configuration: Add the following configuration to the mcpServers section of the JSON file, adjusting the path to server.py to match your local setup:

   { “mcpServers”: { “web-tools”: { “command”: “python”, “args”: [ “/absolute/path/to/MCP/server.py” ] } } }

4. Restart Claude: Restart Claude for Desktop to apply the changes.

5. Verify Integration: You should now see the web_scrape tool available in Claude’s interface. You can ask Claude to fetch content from a website, and it will use the tool.

Example Usage with Claude

Once integrated, you can ask Claude questions like:

• “What’s on the homepage of example.com?”
• “Can you fetch and summarize the content from mozilla.org?”
• “Get the content from wikipedia.org/wiki/Model_Context_Protocol and explain it to me.”

Claude will use the web_scrape tool to fetch the content and provide it in its response.

Extending the Server: Adding Your Own Tools

The UBOS MCP Server is designed to be easily extensible. Adding new tools is a simple process:

1. Create a New Tool File: Create a new Python file in the tools/ directory, e.g., tools/new_tool.py.

2. Implement the Tool Function: Implement your tool function, following a similar pattern to the existing tools.

3. Import and Register the Tool: Import your tool in server.py and register it with the MCP server:

   python

   Import your new tool

   from tools.new_tool import new_tool_function

   Register the tool with the MCP server

   @mcp.tool() async def new_tool(param1: str, param2: int) -> str: “”" Description of what your tool does.

   Args:
       param1: Description of param1
       param2: Description of param2
   
   Returns:
       Description of return value
   """
   return await new_tool_function(param1, param2)
   

4. Restart the Server: Restart the server to apply the changes.

The Streamlit UI: A Visual Interface for MCP Servers

The repository includes a Streamlit application that provides a visual interface for connecting to and testing your MCP servers. This can be particularly useful for debugging and verifying the functionality of your tools.

Features of the Streamlit UI:

• Load and Parse Configuration: Loads and parses your Claude for Desktop configuration file.
• Server Overview: Displays all configured MCP servers.
• Tool Connectivity: Connect to any server and view its available tools.
• Tool Testing: Test tools by providing input parameters and viewing results.
• Resource and Prompt Insights: See available resources and prompts.

Running the Streamlit UI:

bash streamlit run streamlit_app.py

Troubleshooting Common Issues

• Missing Dependencies: Ensure all dependencies in requirements.txt are installed.
• Connection Issues: Verify that the server is running and the configuration in Claude for Desktop points to the correct path.
• Tool Execution Errors: Check the server output for error messages.
• Streamlit UI Issues: Ensure Streamlit is properly installed and the configuration file path is correct.

UBOS: Empowering AI Agent Development

The UBOS MCP Server is a valuable component of the broader UBOS platform, which is designed to be a full-stack AI Agent Development Platform. UBOS focuses on bringing AI Agents to every business department.

Our platform helps you:

• Orchestrate AI Agents: Design and manage complex AI Agent workflows.
• Connect with Enterprise Data: Seamlessly integrate AI Agents with your existing data sources.
• Build Custom AI Agents: Create tailored AI Agents using your own LLM models.
• Develop Multi-Agent Systems: Build sophisticated AI systems that leverage multiple interacting agents.

By providing a powerful and flexible platform for AI Agent development, UBOS is helping businesses to unlock the full potential of AI and transform the way they operate.

Conclusion

The UBOS MCP Server simplifies the integration of web-based tools with LLMs, unlocking new possibilities for AI-powered applications. Whether you’re building a research tool, automating web tasks, or enhancing customer support, the UBOS MCP Server provides a solid foundation for your projects. Combined with the broader capabilities of the UBOS platform, you can create sophisticated AI Agent systems that drive innovation and efficiency within your organization. By embracing the power of MCP and the UBOS platform, you can stay ahead of the curve and leverage the full potential of AI.