Study traditional Chinese divination systems like Bazi and Ziwei Doushu as cultural or historical artifacts.
Generate a structured JSON chart from a birth date and location to feed into an AI agent or analysis pipeline.
Use as a callable skill inside an AI assistant that supports tool-calling to run divination analysis mid-conversation.
Explore how classic Chinese calendrical systems convert between Gregorian and lunisolar dates programmatically.
Install the single dependency (lunar-python via pip), then run the main script with birth date, time, sex, location, and a mode flag. No internet, database, or API key required.
Metaphysics Steward is a Python command-line tool that implements seven systems of traditional Chinese divination and astrology. The README is written in Chinese, and the project is aimed at users interested in studying these practices as cultural artifacts or technical curiosities. The disclaimer at the top of the README explicitly says the tool is for research and cultural learning only, and warns against using it for major life or financial decisions. The seven systems it covers are Bazi (Four Pillars of Destiny, a form of birth-chart analysis based on the year, month, day, and hour of birth), Ziwei Doushu (Purple Star Astrology, a detailed chart system), Qimen Dunjia (a strategic divination system involving a nine-palace grid), Da Liuren (another time-based divination grid method), Meihua Yishu (Plum Blossom Numerology, which uses numbers or the current time to derive a hexagram reading), Jinkoujue (a method attributed to historical strategists), and Liuyao (six-line hexagram divination based on the I Ching). All seven are accessible through a single Python script. You pass in a birth date and time, a sex parameter, and a location (either a city name or a longitude coordinate), along with a mode flag that selects which system to run. The tool corrects for true solar time, which differs from standard clock time depending on longitude. It includes coordinates for over 30 Chinese cities so users can supply a city name rather than a raw longitude. The tool can also output results as JSON, which makes it easier to pipe the structured data into other programs or AI agents. The README describes how AI agents that support tool-calling can load this project as a skill and invoke the analysis directly from a conversation. The only external dependency is a library called lunar-python, which handles conversions between the Gregorian calendar and the traditional Chinese lunisolar calendar. No database, no web service, and no internet connection are required after installation.
← superzhang21 on gitmyhub — every repo by this author, as a profile.
Verify against the repo before relying on details.