Monthly Reading: Calculation Method and Evidence
How OpenFate compares Bazi, Ziwei and Western transits for one month. Check calendar boundaries, synthetic calculation records, report safeguards and remaining evidence gaps.
What is being calculated?
A monthly reading starts from a saved birth date, exact time and place. The selected civil month and its adjacent months are calculated separately; your question guides which facts are discussed. AI writes the interpretation after calculation.
Choose one of five months starting this month, then your topic and saved person. Each new report requires active membership and 65 points; saved reports remain readable after expiry. Each new generation saves a separate report. Retrying a completed request reuses its result without charging again.
A saved report preserves its generated text and calculation evidence. Reading it again does not recalculate the chart with a newer engine. It is a monthly interpretation, not a day-by-day forecast or event calendar.
Why the three systems change at different times
Bazi solar-term months
Month pillars change at the exact Jie solar-term instant, calculated on the Chinese standard calendar and displayed in the report timezone. They do not change automatically on the first day of a civil month.
Ziwei lunar months
Monthly palaces and transformations follow the existing Ziwei lunar-calendar and leap-month rules using the report civil clock. A lunar-month change need not coincide with the Bazi solar term.
Western transit windows
Selected transiting planets are compared with natal targets. Numerical searches estimate when a contact enters its orb, is closest and leaves. A window may overlap either Chinese-calendar change; an estimated closest time is not an appointment for a real event.
What the report checks can catch
Source notes connect chart-derived statements to a system and an interval. Reader-supplied circumstances and suggested actions must remain distinct from calculated facts.
Automatic checks record source, timing, language and editorial observations. These observations do not reject or rewrite the interpretation. Empty or provider-truncated output is retried within the configured limit.
These checks reduce certain errors. Valid source references do not establish that every sentence is supported, and an AI reviewer is not an independent expert. Disagreement between systems should remain visible instead of becoming a majority-vote prediction.
A reproducible calculation record
The downloadable record uses fictional inputs and the local monthly engine. It records the settings, selected observed boundaries and explicit check results. A replay compares the current output with the saved record so a changed result must be investigated.
This is an OpenFate first-party regression record. It does not compare a complete chart against an independent service, test generated report quality or measure real-world prediction accuracy. Individual check groups are not independent people or customer outcomes.
One observed calendar comparison
October 2026, one synthetic birth input, with a midnight day boundary and true solar time disabled. Times from the record are displayed in Asia/Taipei. The archived writing sample uses a different input and a 23:00 day boundary; these are not a same-settings comparison. Calendar changes are not predicted events.
- Bazi solar-term months
- Ziwei lunar months
Open the calculation record (JSON) ·
Replay with the matching source checkout
node --import tsx scripts/verify-monthly-public-evidence.tsRead the actual archived writing
The product page contains a complete Traditional Chinese development report from September 22, 2026, for a fictional Taipei birth record and October 2026 question. It preserves the original prose and source notes. It is a historical example, separate from this calculation replay, and not a guarantee of current report quality.
What has not been established
- A same-settings independent audit of the complete three-system monthly calculation is not established by this record.
- Independent reader blind reviews, consistent value for repeated paid use and prospective prediction accuracy remain unverified.
- The transit selection excludes the transiting Moon and standalone generic lunations. It is not an exhaustive calendar of every ingress, station or possible event.
- Unknown or guessed birth times can change chart-dependent results. Agreement between symbolic systems is not an event probability or proof that the report is the most accurate.
Report a calculation or explanation problem
Contact OpenFate with the target month, report timezone, relevant source label and a short description of the discrepancy. For comparisons, include the other method’s calendar and settings so differences can be evaluated fairly.
Use private support for report issues. Do not post full birth records, another person’s details or an entire private report publicly.