🚀 MCP Server for Document Processing

What is 🚀 MCP Server for Document Processing?

This MCP server allows AI assistants to query and retrieve information from custom document collections, overcoming knowledge cutoffs by processing Markdown, text, PDF, and Word files into vector embeddings stored in a local database. It is designed for developers who want to extend LLM knowledge with up-to-date framework documentation, private codebases, or technical specifications.

How to use 🚀 MCP Server for Document Processing?

Clone the repository, copy .env.example to .env and configure desired settings, then place your Markdown and text files in the data/ directory. Run the pipeline with docker-compose build pipeline && docker-compose run pipeline, then build the server with docker-compose build server. Finally, generate an MCP configuration using the platform‑specific setup script (setup-mcpServer-json.sh or setup-mcpServer-json.bat) and add it to an MCP‑compatible AI assistant such as Roo.

Key features of 🚀 MCP Server for Document Processing

• Processes .md, .txt, .pdf, .docx, and .doc files.
• Supports free local embedding models (e.g., all‑MiniLM‑L6‑v2) and paid OpenAI models.
• Exposes MCP tools: read_md_files, search_content, get_context, project_structure, suggest_implementation.
• Operates in Full Processing Mode (with Claude) or Context Retrieval Mode.
• Fully containerized with Docker for simple setup and portability.
• Customizable chunk size, overlap, batch size, and supported extensions.

Use cases of 🚀 MCP Server for Document Processing

• Provide AI assistants with the latest React 19, Angular 17, or Vue 3.4+ documentation not in training data.
• Enable debugging and understanding of private codebases by feeding proprietary API documentation.
• Import technical specifications or new protocol docs for context‑aware AI assistance.
• Build a searchable knowledge base from internal wikis or blog posts for team use.

FAQ from 🚀 MCP Server for Document Processing

What file types are supported?

By default, the server supports Markdown (.md), Text (.txt), PDF (.pdf), and Word (.docx, .doc) files. You can add more extensions via the SUPPORTED_EXTENSIONS environment variable.

Do I need an API key to run the server?

No. The server can use free local embedding models (e.g., sentence-transformers/all-MiniLM-L6-v2) without any API key. An OpenAI API key is only required if you choose a paid embedding model. An Anthropic API key is optional and enables Full Processing Mode with Claude.

How do I configure the server?

Copy .env.example to .env and edit the environment variables. Key settings include chunk size, embedding model, data directories, and whether to use the Anthropic API. After configuration, run the processing pipeline and then build the server.

What are the two operational modes?

In Full Processing Mode (when `