cafe-mcp

cafe-mcp

Korean Saju (Four Pillars / 사주명리) birth-chart analysis — 5 free tools: chart, five elements, gyeokguk, yin-yang/climate balance, and classical yongshin candidates. Local stdio server via npx, no API key required.

Category
Visit Server

README

cafe-mcp

<!-- mcp-name: io.github.dangamsoft/cafe-mcp -->

CAFE = Cross-weighted Analysis of the Five Elements

Public OWL ontology for Korean Saju (사주명리학), with planned MCP integration via the CAFE engine.

MCP License W3C OWL 2 Ontology Version

v0.6.0 — The MCP server is live. A local (stdio) server exposing 5 free birth-chart toolssaju_chart, ohaeng_balance, gyeokguk, eumyang_johu, yongshin_candidates — alongside the public W3C OWL ontology. See Install below.


What this is

A formal W3C OWL 2 ontology for Korean Saju (사주명리학), grounded in classical Myeongli theory and six canonical treatises. Released under CC BY 4.0 for academic use, system integration, and downstream development.

The ontology is the public layer of the CAFE engine — a deterministic Saju analysis system combining classical theory with ML-based yongshin classification.

W3C OWL 2 온톨로지 형태로 공개된 한국 사주명리학 도메인 모델입니다. 6대 고전(적천수·자평진전·난강망·궁통보감·연해자평·삼명통회)에 근거하며, CC BY 4.0으로 학술·시스템 통합·후속 개발에 자유롭게 사용할 수 있습니다.

이 온톨로지는 CAFE 엔진의 공개 레이어입니다 — 결정적(deterministic) 사주 분석 시스템으로, 명리학 고전 이론과 ML 기반 용신 분류를 결합합니다.


Install (MCP server)

The server runs locally over stdio and talks to the hosted CAFE backend — no API key, no local engine, no database required.

Claude Desktop — add to claude_desktop_config.json:

{
  "mcpServers": {
    "cafe-mcp": {
      "command": "npx",
      "args": ["-y", "@dangamsoft/cafe-mcp"]
    }
  }
}

Cursor — add the same mcpServers entry to ~/.cursor/mcp.json.

Then ask, e.g. "Show my saju for 1990-01-15 10:30, male."

Tools (5, free)

Tool Returns
saju_chart Four Pillars chart — stems/branches, Ten Gods, Spirit Stars
ohaeng_balance Five Elements (五行) distribution
gyeokguk Chart structure/pattern (格局)
eumyang_johu Yin-Yang & Climate (陰陽·調候) balance
yongshin_candidates Classical 5-type Yongshin (用神) candidates

All tools take birth (YYYYMMDDHHMM) and gender (0 = female, 1 = male); name and is_lunar are optional.

The AI-selected final Yongshin (91.1%-accuracy classifier), full Yongshin report, NCODE personality profile and AI naming are part of the 24Plus service and are not exposed here.

Custom backend: set CAFE_MCP_API_URL (the server POSTs to ${CAFE_MCP_API_URL}/try/panels; default https://24plus.ai.kr/api).


Ontology

  • 1,711 RDF triples · SPARQL-queryable · W3C OWL 2 conformant
  • Models:
    • 60갑자 (Sexagenary cycle)
    • 75 sinsal (special-star annotations)
    • 18 patterns — 정격 10 + 특수격 8
    • 5 yongshin types — 격국·억부·병약·조후·통관
    • 12 life stages (12운성)
  • Grounded in 6 classical treatises:
    • 적천수 (Diqianshui)
    • 자평진전 (Zipingzhenquan)
    • 난강망 (Lanjiangwang)
    • 궁통보감 (Qiongtongbaojian)
    • 연해자평 (Yuanhaiziping)
    • 삼명통회 (Sanmingtonghui)

See ontology/ for module structure, SPARQL examples, and the integrated turtle file (cafe-ontology.ttl).


CAFE engine performance

The CAFE engine — the inference system this ontology models — operates inside the 24Plus platform. Validated performance on 5-class yongshin classification (목/화/토/금/수, 5-fold cross-validated):

Metric Result
Accuracy 91.10%
F1-Score (Macro) 91.08%
Precision (Macro) 91.16%
Recall (Macro) 91.10%
ROC AUC (Macro) 97.76%

Random baseline = 20% (5 classes). 4.56× better than random. ML ensemble with weighted voting and probability calibration.

The engine itself remains proprietary; the ontology models the domain structure it operates on.


Note on East Asian Myeongli traditions

This MCP focuses on the Korean Myeongli (사주명리학) tradition, which shares its classical roots — 子平眞詮 (Zipingzhenquan), 滴天髓 (Diqianshui), 窮通寶鑑 (Qiongtongbaojian), 淵海子平 (Yuanhaiziping), 三命通會 (Sanmingtonghui), 欄江網 (Lanjiangwang) — with the Chinese Bazi (八字) tradition.

Romanization uses Korean Revised Romanization (e.g., Yongshin, Sinsal). Hanja (정자) is preserved throughout the ontology.

본 MCP는 한국 명리학(사주명리학) 전통에 기반하며, 중국 八字 전통과 6대 고전(자평진전·적천수·궁통보감·연해자평·삼명통회·난강망)을 공유합니다.


Documentation


Built by

Dangamsoft (단감소프트) — Korean AI company specializing in classical East Asian knowledge systems.

🌐 https://24plus.ai.kr


License

  • Code (when published in v0.6.0): Apache 2.0
  • Public OWL ontology (ontology/): CC BY 4.0

The CAFE inference engine, scoring algorithms, ML model weights, and interpretation templates remain proprietary.


더 깊은 분석(AI 선정 용신 · 전체 리포트)은 https://24plus.ai.kr

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured