oh-my

oh-my-mcp is a Python-based MCP server that exposes a large collection of tools across multiple categories (Compression, Web & Network, File System, Data Processing, Text Processing, System, Utilities, Subagent AI orchestration, and Browser Automation). It is built on the FastMCP framework and provides 116 practical tools designed for modeling context tasks, data manipulation, and automation workflows. Once started, you can connect to the MCP server using a compatible MCP client (e.g., Claude Desktop or other MCP clients) to invoke individual tools, pass parameters, and receive structured results. The server relies on a modular plugin structure, with tools organized under categories such as mcp_server.tools.web, mcp_server.tools.file, mcp_server.tools.data, and more, making it straightforward to discover and compose capabilities for complex tasks.

To use the server, install the package in development mode, configure it via the interactive setup wizard, and then launch the server. The tool suite includes capabilities like web search/backends, HTTP API clients, file operations with safety checks, JSON/CSV/XML/YAML/TOML processing, regex and text extraction, system information, cryptographic utilities, AI subagent orchestration, and Selenium-based browser automation. The server is designed to run standalone without requiring external Python interpreters beyond the runtime you install, and it provides an interactive configuration wizard to tailor tooling and integrations to your use case.

Prerequisites:

• Python 3.12 or higher
• pip (comes with Python)
• Development tools for building/installing a package (optional but recommended for development)

Installation steps:

1. Create and activate a virtual environment (highly recommended):

python -m venv venv
# Windows
venv\Scripts\activate
# macOS/Linux
source venv/bin/activate

2. Install the package in development mode:

pip install -e .

3. Run the interactive configuration wizard (recommended):

uv run configure.py

4. Alternatively, generate Claude Desktop configuration or HTTP config if needed:

python -m mcp_server.cli.config --claude

python -m mcp_server.cli.config --http-server --port 8765

5. Start the MCP server:

python -m mcp_server.main

Notes:

• The project emphasizes standalone operation without needing external Python interpreters beyond the runtime used for this installation.
• See the docs/docs/zh SETUP_GUIDE for detailed steps and wizard usage.

Tips and notes:

• The server exposes a large suite of tools across categories; consult the Documentation/TOOL_REFERENCE for usage examples and parameters.
• Logging configuration is controlled in mcp_server/utils.py; you can adjust log level, outputs, and format.
• The configuration wizard helps tailor tool availability and subagent settings; take advantage of the interactive setup for a smoother initial run.
• Be mindful of safety features: file operation limits, path validation, and safe evaluation rules are in place to protect against unsafe actions.
• If you encounter environment-related issues, ensure your Python PATH and virtual environment are correctly set, and verify that dependencies are installed in the active environment.