Docs/quickstart

Quickstart Guide

Get up and running with TuViMCP in less than 2 minutes. Learn how to compute full 12-palace lá số charts locally and expose them via MCP server.

Requirements

TuViMCP requires Python 3.10+ or Node.js 18+. All calculations execute locally with zero mandatory cloud dependencies.

1Install via PyPI

Install the core server package directly from PyPI:

pip install tuvi-mcp-server

2Python SDK Usage

Import `TuViEngine` directly in your Python application to compute star alignments:

from tuvi_mcp.tuvi_calculator import get_horoscope_chart

# Compute 12-palace matrix (Solar date)
chart = get_horoscope_chart(
    name="Nguyễn Văn A",
    day=24, month=10, year=1985,
    hour_val="14:30",
    gender_val="Nam",
    is_solar=True,
)

# chart = {"thien_ban": {...}, "dia_ban": [...], "cach_cuc": [...]}
# Look up Quan Lộc palace in the 12-house matrix
quan_loc = next(p for p in chart["dia_ban"] if p["cung_chu"] == "Quan lộc")
print(quan_loc["cung_ten"], quan_loc["hanh_cung"])

3Connect via Model Context Protocol (MCP)

Start the stdio MCP server for agentic integration:

tuvi-mcp-server --stdio