Documentation Markdown Converter

🚀 MCP Documentation Server (MCPDoc)

MCPDoc is a documentation server that allows you to load and manage multiple document sources. It supports various ways to specify document sources, such as using configuration files or directly providing URLs.

🚀 Quick Start

📦 Installation

pip install mcpdoc

🚀 Running

By default, the documentation server loads the following content:

1. Load documents from URLs defined in the llms.txt file.
2. Provide a simple command-line interface (CLI) to start the documentation server.

💻 Command-line Usage

mcpdoc [--yaml CONFIG_YAML] [--json CONFIG_JSON] [--urls URL_STRING...] [--follow-redirects] [--timeout SECONDS]

Options:

• --yaml CONFIG_YAML: Use a YAML configuration file to load document sources.
• --json CONFIG_JSON: Use a JSON configuration file to load document sources.
• --urls URL_STRING...: Directly specify the URLs of document sources in the format of name:url or a plain URL.
• --follow-redirects: Enable HTTP redirects (disabled by default).
• --timeout SECONDS: Set the HTTP request timeout (default is 10.0 seconds).

Examples

1. Use a YAML configuration file:

mcpdoc --yaml sample_config.yaml

2. Use a JSON configuration file:

mcpdoc --json sample_config.json

3. Directly specify URLs:

mcpdoc --urls LangGraph:https://langchain-ai.github.io/langgraph/llms.txt

4. Combine and use multiple methods:

mcpdoc --yaml sample_config.yaml --json sample_config.json --follow-redirects --timeout 15

📚 Configuration File Formats

YAML Example (sample_config.yaml)

# Example configuration file for the mcp-mcpdoc server
# Each entry must contain an llms_txt URL and an optional name
- name: LangGraph Python
  llms_txt: https://langchain-ai.github.io/langgraph/llms.txt

JSON Example (sample_config.json)

[
  {
    "name": "LangGraph Python",
    "llms_txt": "https://langchain-ai.github.io/langgraph/llms.txt"
  }
]

💻 Programming Usage

from mcpdoc.main import create_server

# Create a server with document sources
server = create_server(
    [
        {
            "name": "LangGraph Python",
            "llms_txt": "https://langchain-ai.github.io/langgraph/llms.txt",
        },
        # You can add multiple document sources
        # {
        #     "name": "Another Document",
        #     "llms_txt": "https://example.com/llms.txt",
        # },
    ],
    follow_redirects=True,
    timeout=15.0,
)

# Run the server
server.run(transport="stdio")

🔌 Integration with Other Tools

Integration with VS Code

1. Install Claude Code.
2. Run the following command in the terminal to add the MCP server to your project:

claude mcp add-json langgraph-doc '{"type":"stdio","command":"uvx" ,"args":["--from", "mcpdoc", "mcpdoc", "--urls", "langgraph:https://langchain-ai.github.io/langgraph/llms.txt"]}' -s local

3. Update the ~/.claude.json file.
4. Start Claude Code and run the following commands to view the tools:

$ Claude
$ /mcp 

Other Integrations

Integration with MCP Proxy

1. Start the documentation server in the terminal:

uvx serve mcpdoc.cli:main --from mcpdoc

2. Then use the following command in the MCP proxy to load the document source:

llm use mcpdoc -- install https://langchain-ai.github.io/langgraph/llms.txt

✨ Features

• Provide merging and management of multiple document sources.
• Support loading document sources through YAML or JSON configuration files.
• Support directly specifying URLs in the format of name:url or a plain URL.
• Support HTTP redirects and timeout settings.

Example Configurations

YAML Format (sample_config.yaml)

# Example configuration file for the mcp-mcpdoc server
# Each entry must contain an llms_txt URL and an optional name
- name: LangGraph Python
  llms_txt: https://langchain-ai.github.io/langgraph/llms.txt

JSON Format (sample_config.json)

[
  {
    "name": "LangGraph Python",
    "llms_txt": "https://langchain-ai.github.io/langgraph/llms.txt"
  }
]

Command-line Examples

1. Use a YAML configuration file:

mcpdoc --yaml sample_config.yaml