ClimbMCP
# Climbing MCP Server
An MCP (Model Context Protocol) server for serving climbing book PDF chapters to Claude.
## Features
- Store PDF chapters in the `chapters/` directory
- List all available chapters with metadata
- Search for relevant chapters based on query
- Serve PDF content directly to Claude for reading
## Setup
1. Install dependencies:
```bash
npm install
```
2. Add your PDF files to the `chapters/` directory
3. Create corresponding `.json` metadata files for each PDF with:
- `title`: Chapter title
- `description`: Brief description
- `keywords`: Array of relevant keywords for search
## Usage
Add to your Claude Desktop configuration:
```json
{
"mcpServers": {
"climbing-pdf": {
"command": "node",
"args": ["E:/Projects/climbing-mcp-server/src/index.js"]
}
}
}
```
## Available Tools
- `list_chapters`: Lists all available PDF chapters
- `get_chapter`: Retrieves a specific PDF for Claude to read
- `search_relevant_chapter`: Finds the most relevant chapters based on a query
## Adding PDFs
1. Place PDF files in the `chapters/` directory
2. Create a corresponding `.json` file with the same name (e.g., `chapter1.pdf` → `chapter1.json`)
3. The metadata helps with search and relevance scoringTDQS
Scored across 5 tools
Most tools have clear, distinct purposes: listing, searching, section retrieval, extraction, and full-text fetching. The only potential confusion is between extract_chapter_content and get_chapter_text, both of which involve extraction, but their descriptions sufficiently separate the low-level caching action from the combined extraction-and-cleaning result.
All tool names follow a consistent snake_case verb_noun pattern, with clear verbs like list, search, get, and extract. The naming makes the tool set predictable and easy to navigate.
Five tools is a well-scoped size for this server's purpose. Each tool addresses a distinct part of the workflow: discovering books, extracting content, searching text, and retrieving specific sections or readable chapter text.
The tool surface covers the full reading and search lifecycle for a PDF chapter repository: list available books, extract content, search across text, and retrieve sections or readable text. No critical actions are missing for the apparent domain.