A structured timing-analysis protocol derived from Zi Ping / Four Pillars methodology. It translates a classical Chinese timing-analysis system into an AI-executable reasoning chain with fixed steps, conditional outputs, and falsifiable checks.
The protocol does not predict the future. It identifies conditions under which a decision may function as expansion, escape, mismatch, or leverage — and requires those conditions to be stated as testable hypotheses.
What this is for: decision timing, strategic self-audit, structured reasoning, and AI-assisted analysis.
What this is not for: fortune-telling, deterministic prediction, reassurance, or professional advice.
Step 1 — Get your chart
Go to https://www.mingli.info/bazi
Enter your Gregorian birth date, exact birth time (24h), birth city, and gender. Take a screenshot of the chart result.
Step 2 — Load the protocol
Open the file protocol-v2.1.md in this repository. Click the Raw button, select all, and copy the full text. Send it as the first message to Claude or ChatGPT. The AI will confirm the protocol is loaded.
Step 3 — Ask your question
In a second message, attach your chart screenshot and send the prompt from episode-01-prompt.md. The AI will run the full reasoning chain and answer the 6-point framework.
- protocol-v2.1.md — the canonical protocol text. Locked. Do not modify.
- episode-01-prompt.md — the actual prompt used in Episode 01.
- valid-output-checklist.md — pass/fail checklist for advanced users who want to audit AI output against the protocol.
- disclaimer.md — scope and reliance limits.
- changelog.md — version history.
- LICENSE — CC BY 4.0.
If you reference this protocol in writing, video, or research, cite it as:
Time Intelligence Protocol v2.1. https://github.com/timeintelligence/time-intelligence-protocol
If you test the protocol and collect pass/fail evidence, keep the evidence dated. Single-chart failures do not invalidate the protocol. Systemic patterns are handled through the falsification conditions stated in the canonical protocol file.