Pinecone Economic Books

# Quick Start Guide Get your Pinecone Economic Books MCP Server up and running in 5 minutes. ## Step 1: Install Dependencies **Important:** This server requires Python 3.10 or higher. ```bash # Install with Python 3.10 /opt/homebrew/bin/python3.10 -m pip install -r requirements.txt # Or if python3.10 is in your PATH: python3.10 -m pip install -r requirements.txt ``` ## Step 2: Configure Environment Create a `.env` file with your credentials: ```bash cp .env.example .env ``` Edit `.env` and add your Pinecone API key: ```env PINECONE_API_KEY=your-actual-api-key-here PINECONE_INDEX_NAME=economic-books ``` ## Step 3: Test the Server Run the server directly to ensure it works: ```bash /opt/homebrew/bin/python3.10 server.py ``` If you see no errors, the server is ready! ## Step 4: Configure Claude ### For Claude Code Add to `~/.claude.json`: ```json { "mcpServers": { "pinecone-econ": { "command": "/opt/homebrew/bin/python3.10", "args": ["/Users/srulyrosenblat/Developer/pinecone-econ-mcp/server.py"] } } } ``` **Note:** Adjust the path if your python3.10 is located elsewhere (use `which python3.10` to find it). ### For Claude Desktop Add to `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS): ```json { "mcpServers": { "pinecone-econ": { "command": "/opt/homebrew/bin/python3.10", "args": ["/Users/srulyrosenblat/Developer/pinecone-econ-mcp/server.py"] } } } ``` ## Step 5: Restart Claude Restart Claude Code or Claude Desktop to load the MCP server. ## Quick Test Try these searches in Claude: 1. **Semantic search (recommended):** "Use the pinecone-econ MCP server to search for content about market equilibrium and price discovery" 2. **Semantic search with filter:** "Search for labor productivity concepts in books by Wassily Leontief" 3. **Search by author:** "Find all books by Wassily Leontief" 4. **Search by subject:** "Find economic content tagged with the subject 'equilibrium'" 5. **Get index stats:** "Show me the statistics for the economic books index" ## Available Tools The server exposes 10 search tools. **All search tools use semantic search** - no dummy vectors or keyword matching. ### Primary Tools (Recommended) 1. `semantic_search` - Natural language search (SIMPLEST) 2. `semantic_search_with_filters` - Semantic search + metadata filters ### Filtered Semantic Search All tools combine semantic search with metadata filtering: 3. `search_by_author` - Semantic search within author's works 4. `search_by_subject` - Semantic search within tagged topics 5. `search_by_book` - Semantic search within a specific book 6. `search_by_page_range` - Semantic search within page ranges 7. `advanced_search` - Semantic search with multiple filters ### Utility Tools 8. `get_by_id` - Get specific document by ID 9. `get_index_stats` - Get index statistics 10. `vector_search` - Search with pre-computed vectors (advanced) ## Default Namespace All tools use `"book_data"` as the default namespace. You can override this by passing a `namespace` parameter to any tool. ## Troubleshooting **Server doesn't start:** - Check that all dependencies are installed: `pip install -r requirements.txt` - Verify your `.env` file exists and has valid credentials - Ensure Python 3.10+ is being used: `python --version` **No results from searches:** - Verify your Pinecone index name is correct - Check that data exists in the `book_data` namespace - Use `get_index_stats()` to verify index contents **Claude doesn't see the server:** - Ensure you used the absolute path in the config file - Restart Claude completely after adding the MCP server - Check for JSON syntax errors in the config file ## Next Steps - See [README.md](README.md) for detailed usage examples - Review [server.py](server.py) to understand the implementation - Customize the tools for your specific use case ## Support For issues or questions: - Check the [README.md](README.md) for detailed documentation - Review Pinecone logs for API errors - Verify your Pinecone API key has proper permissions