Augmentorium - Ambient Flare

Overview

Augmentorium is a powerful, modular, and extensible Retrieval-Augmented Generation (RAG) platform designed for AI-driven code understanding, project analysis, and knowledge graph workflows. It seamlessly integrates code indexing, context-aware search, and advanced AI agents using the Model Context Protocol (MCP).

Codebase semantic search and RAG
AI agent orchestration (Claude, Klein Claude, GPT, etc.)
Knowledge graph generation and visualization
Multi-project and multi-modal workflows

Features

⚡ Fast, concurrent code indexing

Efficiently indexes large codebases for rapid search and retrieval.

🔎 Semantic search over code and docs

Advanced search capabilities that understand code context.

🧠 Knowledge graph construction

Builds a rich graph representation of projects and their relationships.

🤖 MCP server integration

Communicates via standard IO for seamless agent workflows.

🌐 Web-based graph frontend

Interactive visualizations powered by force-graph.

🔌 Pluggable project support

Add and monitor multiple projects with custom connectors.

🛡️ Strict typing & error reporting

Reliable operation with clear feedback and logs.

📦 Easy deployment

One-line installers & Supervisor config for quick setup.

Demo

"Augmentorium in action: codebase analysis, graph generation, and agent queries."

Installation

Prerequisites

Python 3.8+
Ollama for embeddings: Install
Node.js for MCP server: Install
Git
Supervisor for process management
Optional: npm/Yarn for frontend builds

Clone & Setup

git clone https://github.com/wangdangel/Augmentorium.git
cd Augmentorium
# Linux/Mac
docker-compose up -d
# Windows: run setup_augmentorium_windows.bat as Admin

Usage

All services are managed by Supervisor. To start or restart:

supervisorctl reload

MCP Server Setup

Configure in mcp_config.json:

{
  "augment": {
    "transports": ["stdio"],
    "command": "node",
    "args": ["c:/path/to/Augmentorium/mcp/dist/mcp-server.js"],
    "cwd": "c:/path/to/Augmentorium/mcp"
  }
}

Ollama Embedding Requirements

Install Ollama locally: https://ollama.com/. Embeddings are auto-detected.

Project Structure

Augmentorium/
├── app.py
├── config.yaml
├── requirements.txt
├── frontend/
├── indexer/
├── mcp/
├── scripts/
├── server/
├── utils/
├── tests/
├── setup_augmentorium_*.sh
└── supervisord.conf

Troubleshooting & FAQ

Q: The graph fails to render?
A: Ensure backend services running, restart Supervisor.

Q: Embeddings not found?
A: Verify Ollama installation.

Contributing

Contributions welcome! Open issues or PRs in the repo.

License

Licensed under PolyForm Noncommercial 1.0.0. See license.

Acknowledgments

☄️ Made with passion by AmbientFlare ☄️

☄️ Augmentorium ☄️