QuantConnect PDF MCP Server is an MCP server that indexes PDF collections into searchable markdown for context-aware retrieval across Model Context Protocol clients like Claude Desktop. Originally configured for technical docs like QuantConnect documentation, it connects any directory of PDF files to an underlying TF-IDF search engine with proximity boosting and keyword weighting. Developers, algorithmic traders, researchers, and technical writers use it to explore complex reference materials directly from their conversational interface. The server transforms incoming PDFs into markdown, maps terminology to specific document coordinates, and tracks updates via MD5 hashing to avoid reindexing unchanged files. By exposing dedicated tools for search, file listing, and full-text retrieval, it enables large language models to cite specific pages, review sections, and locate documentation without manual copy-pasting. Users can configure domain-specific terms, parallel processing options, and context snippet limits to tune the retrieval engine for their specific documentation library.
Category: Developer Tools & Code Intelligence
Tags: documentation, finance, markdown, pdf, quantconnect
Visit QuantConnect PDF MCP Server
bash git clone https://github.com/lhstorm/mcp_server_knowledge_engine.git cd mcp_server_knowledge_engine 2. Create and activate a Python virtual environment, then install dependencies: bash python -m venv venv source venv/bin/activate pip install -r requirements.txt 3. Initialize your configuration by running python manage_server.py create-config and specifying your PDF directory and domain keywords. 4. Add and process your documents: bash python manage_server.py process-pdfs 5. Automatically merge the configuration into Claude Desktop by running: bash python generate_mcp_config.py --merge 6. Restart Claude Desktop to start using the tools.Part of MCP Servers
Clone the repository from GitHub, create a Python 3.8 or higher virtual environment, and install the required packages using pip install -r requirements.txt. Afterward, run python manage_server.py create-config to establish your server settings, process your PDFs using python manage_server.py process-pdfs, and merge the configuration into Claude Desktop using python generate_mcp_config.py --merge.
The server provides three configurable tools: a search tool (default search_docs) that conducts TF-IDF proximity search across indexed markdown; a list tool (default list_docs) that returns metadata and page counts for all available PDFs; and a content tool (default get_document_content) that fetches markdown text or specific page ranges from documents.
It is designed to work with Claude Desktop, VS Code, and any other MCP-compliant client. The repository includes an automated utility to generate and merge configuration files directly for Claude Desktop environments.
Yes. While named after QuantConnect PDF documentation, the underlying engine handles universal PDF collections. You can configure custom domain keywords and directory paths to index legal papers, medical records, research papers, or any other technical documentation.