Provides web search capabilities through Brave Search API
Provides containerized deployment of the MCP server
Used for cloning the repository and version control
Used as the implementation language for the MCP server and custom tools
Mentioned as a data source that can be accessed through the MCP toolkit
Mentioned as a platform that can be controlled through the MCP toolkit
Provides a web-based configuration UI for the MCP toolkit
Referenced in example queries for stock data analysis through Yahoo Finance integration
MCP Tool Kit
A modular server implementation for building high precision vertical AI agents. Intended for use building high precision vertical AI agents, but can deploy to gain access to general tool functionality.
Uses >=50% less code than the Python MCP SDK alone.
Benefits of This Approach
Reduces Cognitive Load on Claude: Claude doesn't need to reason about the sequence of tool calls
Encapsulates Domain Knowledge: The agent can contain domain-specific logic about how to perform tasks well in a particular vertical
Simplifies Error Handling: The agent can handle errors and retries internally without Claude's involvement
Enables Complex Workflows: Multi-step processes that would be difficult to coordinate through individual tool calls
Maintains Conversational Flow: The user isn't exposed to the complexity of the underlying system
Example Scenario
Here's a concrete example of Claude invoking a vertical agent:
Overview
The MCP Unified Server provides a unified interface for Claude to interact with various external systems and tools including:
File system operations: Read, write, and manipulate files
Time tools: Get current time in different timezones, convert between timezones
Sequential thinking: A tool for dynamic and reflective problem-solving
Brave Search: Web and local search capabilities
Browser automation: Complete browser control via Playwright
World Bank API: Access to economic and development data
News API: Access to global news sources and articles
PowerPoint: Create and manipulate PowerPoint presentations
Excel: Create and manipulate Excel spreadsheets
Yahoo Finance: Stock market and financial data
FRED: Federal Reserve Economic Data
Agentic capabilities: Create and deploy autonomous agents that perform complex tasks
And many more specialized tools
123 Total Tools Available
Quickstart Guide: Deploy Your First MCP Server with Default Tools
Note: ensure that you have git downloaded (https://git-scm.com/downloads) and Docker downloaded (https://www.docker.com/products/docker-desktop/) and running. You also must ensure the git executable file is added to path (instructions towards end of this README).
Docker deployment (recommended & most stable)
Clone the repository:
You can then use Docker in one of two ways: Option 1 - Using docker-compose:
Option 2 - Direct Docker command:
The repository includes a sample Claude desktop configuration file (claude_desktop_config.json
) that you can use:
Troubleshooting docker
If you are getting errors running docker, it is likely that the image name is incorrect in the Claude desktop configuration file. A common fix is to use the following json for configuration:
Install via pip
Local server basic configuration:
Configure Claude Desktop to Access Your Server
Open Claude Desktop app
Go to File > Settings > Developer > Edit config
Add the 'claude_desktop_configuration.json' file
Save the configuration
Restart the MCP server with your new tool integrated
Restart and Open Claude Desktop app (for windows users you must use task manager to end task for all Claude instances)
You can import this configuration in the Claude desktop app or use it as a reference to create your own.
You now have immediate access to powerful capabilities including file operations, web search, time tools, and more—without requiring any API keys or complex setup.
Setting Up Environment Variables
After cloning the repository, you have two options to configure your environment variables:
Option 1: Interactive Setup Script
Run the setup script which will guide you through setting up your environment variables:
This script will create a .env
file in the repository with your configuration.
Option 2: Manual Configuration
Alternatively, you can manually create a .env
file in the repository root with the following variables:
Sample Claude Prompts
Once set up, you can ask Claude to use the tools with prompts like:
"Search the web for the latest AI research papers and summarize the findings."
"Create a PowerPoint presentation about climate change with three slides."
"Use the weather_checker agent to tell me the current conditions in Tokyo."
"Can you use the quick_lookup agent to research quantum computing advances?"
"Download my QuickBooks invoice data and analyze our revenue for the past quarter."
"Set up a product on my Shopify store with these details and pricing."
"Get the current stock price and historical data for Tesla using Yahoo Finance."
"Analyze inflation trends using FRED economic data for the past 5 years."
"Use browser automation to fill out this form at [website URL]."
"Read the text file in my Downloads folder named 'project_notes.txt'."
"Get the latest news headlines about technology."
Available Tools
File System Tools
read_file
: Read contents of a fileread_multiple_files
: Read multiple files simultaneouslywrite_file
: Create or overwrite a fileedit_file
: Make line-based edits to a filecreate_directory
: Create a new directorylist_directory
: Get directory contentsdirectory_tree
: Get a recursive tree viewmove_file
: Move or rename files/directoriessearch_files
: Search for files matching a patternget_file_info
: Get file metadatalist_allowed_directories
: List allowed directoriesBrowser_Automation:
playwright_launch_browser
: Launch a new browser instanceplaywright_navigate
: Navigate to a URLplaywright_screenshot
: Take a screenshotplaywright_click
: Click on an elementplaywright_fill
: Fill an input fieldplaywright_evaluate
: Execute JavaScriptplaywright_get_content
: Get the HTML content of a page
Agent Tools
run_agent
: Execute a registered agent with parameterslist_agents
: List all available agents and their metadata
Financial Data Tools
Yahoo Finance:
yfinance
: Get stock quotes and historical datayfinance_get_quote
: Get current stock quoteyfinance_get_history
: Get historical stock datayfinance_get_info
: Get detailed company informationyfinance_get_options
: Get options chain datayfinance_get_recommendations
: Get analyst recommendations
FRED (Federal Reserve Economic Data):
fred_get_series
: Get economic data seriesfred_get_series_info
: Get metadata about a seriesfred_search
: Search for economic data seriesfred_get_category
: Browse data by categoryfred_get_releases
: Get economic data releasesfred_get_sources
: Get data sources
Time Tools
get_current_time
: Get current time in a specified timezoneconvert_time
: Convert time between timezones
Sequential Thinking
sequentialthinking
: A tool for breaking down complex problems using a step-by-step thinking process
Brave Search
brave_web_search
: Perform web searchesbrave_local_search
: Search for local businesses and places
World Bank API
worldbank_get_indicator
: Get indicator data for a country
News API
news_top_headlines
: Get top news headlinesnews_search
: Search for news articlesnews_sources
: List available news sources
PowerPoint Tools
ppt_create_presentation
: Create a new PowerPoint presentationppt_open_presentation
: Open an existing presentationppt_save_presentation
: Save a presentationppt_add_slide
: Add a new slideppt_add_text
: Add text to a slideppt_add_image
: Add an image to a slideppt_add_chart
: Add a chart to a slideppt_add_table
: Add a table to a slideppt_analyze_presentation
: Analyze presentation structureppt_enhance_presentation
: Suggest enhancementsppt_generate_presentation
: Generate a presentation from textppt_command
: Process natural language commands
For a complete list of available tools, see the documentation or browse the tools directory. For a complete list of available tools, see the documentation or browse the tools directory.
Adding a New Tool Module
Create a new file in the
tools
directory (e.g.,my_tool.py
)Follow the existing module pattern:
Create service class
Define tool functions
Implement registration functions
Update
mcp_unified_server.py
to import and register your new module
Extending an Existing Tool Module
Add new methods to the service class
Add new tool functions
Update the registration function to include your new tools
Development with Docker
You can use Docker for development to ensure a consistent environment:
This mounts your local repository into the container, so changes to the code are reflected immediately (for most files).
Philosophical Perspective: The Human-AI Cognitive Partnership
The MCP Tool Kit represents a paradigm shift in how we conceptualize the relationship between human intelligence and AI systems. Rather than positioning AI as a mere tool for task automation, this framework establishes a cognitive partnership where human strategic thinking and AI operational capabilities complement each other in profound ways.
The agentic architecture embodies a transformative vision: AI systems that can independently interpret context, make decisions within bounded parameters, and execute complex sequences of actions—all while maintaining human oversight and strategic direction. This represents not merely a technological advance, but a fundamentally new model for human-machine collaboration.
In this evolving cognitive landscape, the most successful implementations will be those that thoughtfully balance technological potential with human capabilities, creating interfaces that enhance rather than replace human decision-making and creativity.
Troubleshooting
Module not loading: Check the import path and dependencies
API key errors: Verify your API keys in the
.env
filePermission errors: Check the allowed directories in
MCP_FILESYSTEM_DIRS
Connection errors: Ensure the server is running and the port is accessible
Agent not detected: Verify the agent file is in the correct directory and follows the required format
Issues with path: Make sure that git is added to path as per the below instruciton
#To add Git to the PATH on Windows, follow these steps:
Locate Git Installation: Determine the path where Git is installed on your system. Common paths include C:\Program Files\Git\bin\git.exe and C:\Program Files\Git\cmd for a standard Git installation, or C:\Users<username>\AppData\Local\GitHub\PortableGit_\bin and C:\Users<username>\AppData\Local\GitHub\PortableGit_\cmd if you installed Git through GitHub for Windows or GitHub Desktop.
Edit Environment Variables: Open the "Edit Environment Variables" app either through the Control Panel or by searching for "Edit the system environment variables" in the Start menu. Under the "System variables" section, find the "Path" variable, click "Edit...", and add the path to the Git executable and command files. Ensure there are no spaces around the semicolons separating paths.
Save Changes: After adding the Git paths, click "OK" to save your changes. Close and reopen any command prompt windows to apply the new PATH settings.
Verify Installation: Open a command prompt and run git --version to verify that Git is accessible from the command line.
License
The MCP Unified Server is licensed under the MIT License.
Acknowledgements
This project uses several open-source libraries and APIs:
MCP SDK for Claude AI assistants
NewsAPI for news access
Brave Search API for web search
World Bank API for economic data
python-pptx for PowerPoint manipulation
XlsxWriter for Excel spreadsheets
This server cannot be installed
hybrid server
The server is able to function both locally and remotely, depending on the configuration or use case.
A modular server implementation for Claude AI assistants with integrated tools, enabling Claude to perform actions and access external resources like file systems, web searches, browser automation, financial data, and document generation.
- Benefits of This Approach
- Example Scenario
- Overview
- 123 Total Tools Available
- Quickstart Guide: Deploy Your First MCP Server with Default Tools
- Docker deployment (recommended & most stable)
- Install via pip
- Configure Claude Desktop to Access Your Server
- Setting Up Environment Variables
- Available Tools
- Philosophical Perspective: The Human-AI Cognitive Partnership
- Troubleshooting
- License
- Acknowledgements
Related Resources
Related MCP Servers
- -securityFlicense-qualityA comprehensive suite of Model Context Protocol servers designed to extend AI agent Claude's capabilities with integrations for knowledge management, reasoning, advanced search, news access, and workspace tools.Last updated -5
- AsecurityFlicenseAqualityThe server connects the Claude AI to the internet through the capabilites of the LSD SQL language, turning web data into a queryable database-like structure to interact with real-world data efficiently.Last updated -464
- -securityFlicense-qualityA smart documentation server that provides AI-assisted code improvement and documentation management through Claude Desktop integration.Last updated -10
LPS MCPofficial
-securityAlicense-qualityA minimal server that provides Claude AI with secure file system access and sequential thinking capabilities, allowing Claude to navigate directories, read files, and break down complex problems into structured thinking steps.Last updated -1MIT License