Library Docs MCP Server
# Library Docs MCP Server
This is an MCP (Model Context Protocol) server that allows you to search and fetch documentation for popular libraries like Langchain, Llama-Index, MCP, and OpenAI using the Serper API.
## Features
- Search library documentation using a natural language query.
- Supports Langchain, Llama-Index, MCP, and OpenAI (Update the code to add other libraries).
- Uses the `Serper API` to perform site-specific searches.
- Parses and returns the documentation using `BeautifulSoup`.
- **Provides updated documentation** – useful for LLM models with knowledge cut-off dates.
## Why Use This Server with LLMs?
Many LLM models, including those used in **Claude Desktop** and similar platforms, have a knowledge cut-off date and may not have access to the latest library documentation. This MCP server solves that problem by:
- Fetching **real-time documentation** from official sources.
- Providing **up-to-date information** for development and troubleshooting.
- Improving the accuracy and relevance of responses when working with new library updates.
## Setting Up with Claude Desktop
To use this server with **Claude Desktop**, update the `claude_desktop_config.json` file with the following configuration:
```json
{
"mcpServers": {
"docs-mcp-server": {
"command": "C:\\Users\\Vikram\\.local\\bin\\uv.exe",
"args": [
"run",
"--with",
"mcp[cli]",
"mcp",
"run",
"F:\\My Projects\\AI\\docs-mcp-server\\server.py"
]
}
}
}
```
TDQS
Scored across 1 tool
With only one tool, there is no possibility of ambiguity or overlap between tools. The single tool 'get_docs' has a clear, distinct purpose of searching documentation for specified libraries.
The single tool name 'get_docs' follows a clear verb_noun pattern. With only one tool, naming consistency is inherently perfect as there are no other tools to compare against.
A single tool is insufficient for a documentation search server. While the tool covers basic search functionality, the server lacks tools for browsing documentation structure, getting library metadata, or handling multiple queries efficiently, making the tool count too low for the apparent scope.
The tool surface is severely incomplete for a documentation server. There are significant gaps: no ability to list available libraries, browse documentation hierarchies, get version-specific docs, or handle pagination/search refinement. The single search tool provides only basic functionality without supporting common documentation workflows.