🧠 myBrAIn — MCP Ecosystem

Deterministic, Idempotent, Context-Aware Second Brain for Google Antigravity.

---

Introduction

myBrAIn is an MCP (Model Context Protocol) server designed to provide persistent and contextual memory to language models (like Google Antigravity). It acts as a "second brain" for your development environment, allowing the AI to remember project rules, architectural decisions, and technical insights across different chat sessions.

Core Principles

1. Idempotency: Uses a deterministic Hash ID system to avoid duplicates and ensure data consistency.
2. Security: Designed to operate in MCP environments; logs are sent exclusively to Stderr to avoid polluting the Stdout channel used by the protocol.
3. Robustness: Built-in Vector Conflict Detection system to prevent the storage of conflicting information.

---

System Architecture

The system is divided into three main components:

• Core: Manages persistence (db.py) and analytical intelligence (analyzer.py) for extracting insights from code.
• Server: The standard MCP interface (server.py) that allows Antigravity to interact with the brain.
• Admin UI: An intuitive dashboard (admin.py) based on Streamlit to visualize, manage, and debug stored memories.

---

🛠 Installation (Step-by-Step)

1. Clone the repository:

   git clone https://github.com/lilium360/myBrAIn.git
   cd mybrain

2. Create virtual environment:

   • Windows:

     python -m venv venv
     .\venv\Scripts\activate

   • Mac/Linux:

     python3 -m venv venv
     source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

---

🐳 Docker Deployment

The easiest way to run myBrAIn is using Docker Compose. This starts both the Streamlit Admin UI and prepares the MCP server environment.

Quick Start

# 1. Clone the repository (if not already done)
git clone https://github.com/Labontese/myBrAIn.git
cd myBrAIn

# 2. Prepare environment (optional)
cp .env.example .env

# 3. Spin up the ecosystem
docker compose up -d

Accessing the Dashboard

The Admin UI will be available at: http://localhost:8501

Integration in IDEs (Cursor, VS Code + Antigravity)

To use the MCP server from within your IDE while it's running in Docker, add this to your mcpServers configuration:

{
  "mcpServers": {
    "mybrain": {
      "command": "docker",
      "args": ["exec", "-i", "mybrain-mcp", "python", "server.py"],
      "env": {
        "MYBRAIN_DATA_DIR": "/data/mybrain"
      }
    }
  }
}

---

🚀 Project Onboarding & Context

myBrAIn is specifically designed to handle "Project Onboarding" — a process where the AI analyzes your current codebase and stores its architectural DNA, coding patterns, and constraints in its persistent memory.

How to Onboard a New Project

Once you have the MCP server configured in your IDE, simply use the following prompt to let myBrAIn index your project:

"I want to perform a full onboarding of this project into your long-term memory. Sequentially perform: Structural Link, Stack Analysis, Architectural Analysis, Style Analysis, and Exclusion Analysis."

Available Tools

• initialize_workbase: Link a directory to the brain.
• store_insight: Manually save a rule or context.
• recall_context: Retrieve relevant memories for the current task.
• critique_code: Validate code against stored architectural rules.

---

Advanced Configuration

You can customize the system behavior (similarity thresholds, database paths, embedding models) by modifying the file: mybrain/core/config.py

---

Roadmap & Limitations

• Core v1: Currently focused on vector persistence and code analysis.
• Graph Visualization: (Planned) Visualization of relationships between insights.
• Vector-Based Conflict Detection: Uses cosine similarity to identify potentially contradictory insights.