A comprehensive platform integrating traditional Chinese metaphysics with AI, taibu offers high-precision analysis across numerous divination systems. It targets users interested in exploring their destiny through traditional methods enhanced by modern AI, as well as developers seeking to embed these tools via its MCP server. The project aims to make complex metaphysical knowledge accessible and actionable through an AI-powered interface.

How It Works

taibu deeply fuses traditional Chinese metaphysical systems like Bazi, Ziwei Doushu, and Liu Yao with AI. It supports exporting data from these systems for AI analysis, leveraging multi-model capabilities for deep reasoning and even visual recognition. A key feature is its MCP (Model Context Protocol) server, enabling direct integration of its metaphysical tools into compatible client applications. The platform also manages historical records, allows knowledge base integration, and offers AI personalization options for user interaction.

Quick Start & Requirements

• Docker Deployment: Use docker compose up -d --build for a full stack, or specific docker-compose.*.yml files for Web or MCP server only. Default ports are 3000 (Web) and 3001 (MCP).
• Development Environment: Requires Node.js 18+ and pnpm. Clone the repository, install dependencies (pnpm install), configure .env from .env.example, and run pnpm dev.
• Online Demo: https://www.mingai.fun
• Issue Reporting: https://github.com/hhszzzz/taibu/issues

Highlighted Details

• Extensive Metaphysical Coverage: Supports over a dozen systems including Bazi, Ziwei Doushu, Liu Yao, Qimen Dunjia, Taiyi Shenshu, Daliuren, Meihua Yishu, Tarot, MBTI, face/palm reading, and more.
• Advanced AI Integration: Enables AI analysis of metaphysical data, offering deep reasoning, personalized outputs, and knowledge base integration.
• MCP Server: Provides a dedicated MCP server for seamless integration of metaphysical tools into other applications.
• Multi-Platform Experience: Offers Web, iOS, and Android client applications.
• Community Features: Includes features for community interaction, anonymous discussions, and incentive programs.

Maintenance & Community

The project is primarily developed by hhszzzz. Contributions via Issues and Pull Requests are welcomed. The primary language of documentation and community interaction appears to be Chinese, with an English README available.

Licensing & Compatibility

This repository employs a mixed licensing model. The core packages (taibu-core, taibu-mcp, taibu-mcp-server) are distributed under the MIT License, permitting commercial use, modification, and redistribution with attribution. However, the rest of the codebase, including the Web UI, server components, and deployment scripts, is licensed under AGPL-3.0-only. While AGPL-3.0-only does not prohibit commercial use, it mandates that any modified versions offered as network services must also be made available under the AGPL-3.0-only license, requiring source code disclosure.

Limitations & Caveats

The AGPL-3.0-only license imposes significant copyleft obligations on network services, potentially complicating commercial deployment or integration into proprietary systems. The primary language being Chinese may present a barrier for non-Chinese speaking developers or community members. Specific performance metrics or benchmarks for the AI analysis capabilities are not detailed in the README.