Tech Neoforge Knowledge Base Document Ingest + MCP Server

Summary

An MCP Server that provides Retrieval-Augmented Generation (RAG) capabilities to surface the latest Neoforge Mod API documentation (currently v1.21.11). It is designed specifically for local usage with:

• Inference Engines (LM Studio, Ollama UI etc...)
• LLM Harnesses (OpenCode, ClaudeCode etc...)

Docker Services Include:

• MCP Server: built with FastMCP library.
• Vector Database Server: ChromaDB instance with Neoforge Mod API Knowledge Base
• Ingest Pipeline: Python data ingest reading HTML + Markdown documentation and writing to ChromaDB
• LLM Inference Engine Server: Ollama instance using Qwen3-embedding:0.6b embedding model

Works Well with:

• Neoforge OpenCode Skills

Getting Started

• Prequisite Before Starting
• Ingestion Pipeline
• MCP Server
• Setup Tooling

Documentation

• The Purpose of this Repository
• References

License

This code is distributed under the terms and conditions of the MIT License

---

Spudmash Media [-] 2026