MCP Tool Kit

MCP Tool Kit is an MCP server that provides a unified, modular interface for AI agents to interact with over 120 built-in tools covering file management, web search, browser automation, and financial analysis. It connects LLM environments to external systems including Brave Search, Playwright browser instances, and economic data sources like Yahoo Finance, FRED, and the World Bank. Developers and AI practitioners use it to equip conversational clients and automated agents with extensive operational capabilities without configuring standalone tool integrations separately. The server supports dynamic tool discovery and runtime registration, managing all tool permissions through a central configuration file. By isolating filesystem operations to a designated storage folder, it enables secure file creation, manipulation of office documents, and temporary data processing. In addition, MCP Tool Kit offers dual transport compatibility via standard I/O and Server-Sent Events, allowing it to integrate directly with desktop interfaces or serve remote web-based architectures.

Category: Files, Documents & PDFs

Tags: browser automation, file-operations, financial-data, web search

Visit MCP Tool Kit

How to install and configure MCP Tool Kit

1. Run the platform installer script to set up dependencies and Docker environments automatically: - Windows (PowerShell): irm https://raw.githubusercontent.com/getfounded/mcp-tool-kit/main/install.ps1 | iex - macOS: bash <(curl -s https://raw.githubusercontent.com/getfounded/mcp-tool-kit/main/install.sh) - Linux: bash <(wget -qO- https://raw.githubusercontent.com/getfounded/mcp-tool-kit/main/install.sh) 2. Copy .env.template to .env and configure required variables such as BRAVE_SEARCH_API_KEY and FRED_API_KEY. 3. Start the server in stdio mode using the platform launcher script (launch.bat or ./launch.sh). 4. Add the server to your Claude Desktop configuration file: json { "mcpServers": { "mcp-tool-kit": { "command": "docker", "args": ["exec", "-i", "mcp-server", "python", "mcp_server_v2.py"] } } } 5. Restart Claude Desktop to load the tools.

What you can do with MCP Tool Kit

• Automating web research and live data retrieval using integrated Brave Search queries inside Claude Desktop. - Driving headless Playwright browser automation workflows to interact with complex websites and capture data directly. - Reading, writing, and organizing local documents, spreadsheets, and PDFs within a sandboxed storage workspace. - Fetching historical macroeconomic indicators and market statistics from FRED, Yahoo Finance, and World Bank APIs. - Serving modular agent tools over SSE endpoints to connect web-based agents at localhost on port 8080.

Key facts

• https://github.com/getfounded/mcp-tool-kit
• Files, Documents & PDFs, Other & General Purpose, Web Search & Research
• browser automation, file-operations, financial-data, web search

Part of MCP Servers

How do I install MCP Tool Kit?

You can install it using one-line scripts for Windows PowerShell, macOS, or Linux, which set up the application and check for Docker. Alternatively, you can clone the GitHub repository and run the launch script, or pull and execute the prebuilt Docker container from GitHub Container Registry directly.

What can MCP Tool Kit do?

It provides over 120 tools across multiple domains. These include sandboxed file operations, document manipulation for PDF and Excel files, web search via Brave, Playwright browser automation, time zone conversions, and financial data fetching from Yahoo Finance, FRED, and the World Bank.

Which MCP clients work with MCP Tool Kit?

MCP Tool Kit supports standard input and output transport for direct connection with Claude Desktop. It also supports a Server-Sent Events transport mode running on port 8080, which allows custom web clients and external AI agents to connect via dedicated HTTP and event endpoints.

Is MCP Tool Kit open source?

Yes, MCP Tool Kit is open source and licensed under the MIT License. The complete source code, developer guides, and issue tracker are publicly available on GitHub.