Library
Search a concept or aliasSearch the Wiki
OpenFate

How OpenFate Validates Chart Calculations and AI Reports

A transparent, calculation-first account of how OpenFate turns structured input into chart facts, checks the result, uses AI for explanation, and preserves completed reports.

Published by OpenFate · Methodology reviewed

Overview

For chart-based products, OpenFate does not ask a language model to invent the source chart. Structured input first passes through a system-specific calculation engine. The resulting facts are checked and then supplied to AI for explanation. Focused tests and output contracts reduce errors and regressions; they do not guarantee every interpretation or future outcome.

What Is Calculated and What Is Interpreted

Trust starts by separating reproducible chart facts from interpretive narrative. OpenFate treats these as different layers with different validation needs.

Input contract
Birth date, time, place, timezone, calendar choices, and system settings are normalized before calculation. Missing or incorrect input can change every downstream result.
Deterministic calculation
The same normalized input and settings are designed to return the same structured pillars, placements, cycles, or other system facts.
AI-assisted explanation
The calculation result is the source of chart facts. AI receives those facts as evidence and produces narrative explanation; it is not treated as the authoritative source for placements or chart structure.
Saved report
A completed report is stored as a frozen snapshot. Later engine or prompt changes do not silently rewrite the report a user originally received.

The Calculation-First Publication Pipeline

Each product family has its own rules, but chart-based reports follow the same high-level evidence path.

  • 1. Normalize the user input, timezone, location, and calculation options that the selected system requires.
  • 2. Run the system-specific engine to produce structured chart facts before requesting any narrative analysis.
  • 3. Check required fields, value ranges, internal relationships, and report contracts with deterministic validators.
  • 4. Give the validated evidence to the AI layer with instructions to explain the supplied chart instead of inventing a replacement chart.
  • 5. Apply response-schema, publication, redaction, and rendering checks before saving the completed report as a snapshot.

Systems and Features Covered

Validation is organized by system family. OpenFate does not pretend that one universal benchmark can test every chart, relationship, timing, and oracle product in the same way.

Core chart engines
Bazi (Four Pillars), Ziwei Doushu, Western astrology, Vedic astrology, Human Design, and Japanese Sukuyo.
Relationship analysis
Bazi, Ziwei Doushu, Western astrology, Vedic astrology, and Sukuyo compatibility, plus a multi-system Combination report that preserves system-specific evidence instead of forcing one universal score.
Oracle
Tarot, Lenormand, Meihua Yishu, Liuyao, Qimen Dunjia, and Lingqian, each with its own draw, casting, or interpretation contract.
Focused and supporting features
Bazi Wealth and Love reports, BZTI personality, chart-only views for the six core systems, Huangli, calendar conversion, timezone handling, and True Solar Time or birth-time correction where supported.

How We Check for Regressions

The repository uses several forms of evidence because no single test can establish the correctness of an entire metaphysical product.

  • Dedicated test batteries cover Bazi, Ziwei Doushu, Western astrology, Vedic astrology, Human Design, Sukuyo, Huangli, and all six oracle product families.
  • Selected engines use golden fixtures: known input and expected structured output are locked together so an unexplained calculation change is visible.
  • Regression fixtures preserve previously verified edge cases, including calendar boundaries, timezone behavior, interaction rules, and report rendering contracts.
  • Compatibility tests check system-specific evidence and cross-system report assembly without assuming that every tradition should yield the same score.
  • Response schemas and publication checks reject malformed or incomplete AI output before it becomes a finished user-facing report.
  • Hydration and frozen-report checks make sure saved results render from their original evidence instead of being silently recalculated by a newer view.

How to Check Your Own Result

A useful comparison starts with structured facts, not with whether two narrative readings use the same wording.

  • Confirm the birth date, exact or estimated time, birthplace, timezone, daylight-saving status, and calendar format before comparing charts.
  • Record the calculation convention: for example, whether True Solar Time, a day-boundary rule, a house system, or a sidereal setting is enabled.
  • Compare pillars, branches, placements, houses, cycles, or cast results first. Only compare narrative interpretation after the source facts agree.
  • When two products differ, change one input or convention at a time. This usually reveals whether the mismatch is data, settings, school, or an engine defect.
  • If a mismatch remains, contact OpenFate with the route, non-sensitive input summary, settings, and expected structured fact. Do not send account passwords or unnecessary personal data.

What This Method Does Not Prove

Transparency is more credible when its boundaries are explicit.

  • Automated tests reduce known defect and regression risk; they do not prove that software has zero bugs.
  • A reproducible chart does not objectively validate a metaphysical tradition, interpretation, or prediction of future events.
  • AI-generated explanation can still be incomplete, overconfident, or wrong and must not replace medical, legal, financial, or other qualified professional advice.
  • Different sites may legitimately disagree because of timezone data, daylight saving, True Solar Time, day boundaries, ephemerides, house or zodiac settings, and traditional schools.
  • Incorrect or incomplete birth information propagates through a technically correct calculation.
  • This page is OpenFate first-party disclosure, not an independent audit, certification, or comparative accuracy benchmark.

Independently Checkable Evidence

These public destinations let readers inspect the product boundary, reproduce a chart, and review the terminology and data policies behind the methodology.

Common Questions

Does OpenFate guarantee chart accuracy?
No. OpenFate validates calculations and report contracts to reduce error and regression risk; it does not promise zero defects or certify metaphysical predictions.
Does AI calculate the chart?
For chart-based reports, the system-specific engine is the source of structured chart facts. AI receives those facts to produce an explanation.
Why can another site show a different result?
The input or convention may differ, including timezone, daylight saving, True Solar Time, day boundary, zodiac, house system, ephemeris, or traditional school.
Are all OpenFate systems validated in exactly the same way?
No. Shared report contracts cover common boundaries, while each chart and oracle family also needs system-specific fixtures, invariants, and review criteria.
Can an old saved report change after an update?
Completed reports are designed to render from frozen snapshots. A new calculation may reflect deliberate engine or prompt changes, but it should not silently rewrite the original saved report.
Is this an independent audit?
No. It is a first-party description of OpenFate methods. The public Bazi MCP can be inspected independently, but that visibility is not third-party certification of every OpenFate product.