FOR DEVELOPERS
Good data.
Clear intentions.
Explore three local JSON examples and a practical design checklist. No signup, API key, endpoint subscription, or personal information required.
LOCAL EXAMPLE LAB
Inspect a response.
Change the example, copy its JSON, and examine the purpose of each field. Nothing is submitted.
{
"kind": "natal_chart",
"mode": "illustrative_only",
"settings": {
"zodiac": "tropical",
"house_system": "whole_sign"
},
"placements": [],
"warning": "No birth data or calculation supplied."
}Local illustrative data. No API call, birth data, or live calculation. These are proposed response shapes, not hosted endpoint documentation.
Read the response, then the assumptions
These examples are static fixtures for discussing interface design. Switching a tab changes the local JSON shown on the page. It does not call an API, calculate a placement, or use personal information. The schemas are proposed teaching examples, not documentation for an available GeminiAPI.com endpoint.
Three examples, three purposes
The natal example names settings and leaves placements empty because no birth data or calculation is supplied. The horoscope example contains a general editorial reflection rather than a personal chart result. The relationship example leaves comparisons empty and does not invent a compatibility percentage.
Start with the example closest to your intended experience. Decide which fields are supplied facts, which are calculation settings, which are generated results, and which are editorial content. A clear distinction makes errors easier to explain and assumptions harder to overlook.
Field notes
| Field | Purpose | Design boundary |
|---|---|---|
kind | Names the example’s output type. | It does not imply a live service. |
mode | Identifies an illustrative or editorial example. | Keep it visible near the displayed content. |
settings | Records selected conventions. | Settings are not biographical facts. |
placements | Would contain calculated data in a real implementation. | An empty array is not a failed personal reading. |
warning | Explains the scope of a result. | Do not hide essential limits in a distant footer. |
theme | Holds an editorial reflection. | It is not a prediction or assessment. |
A sensible integration plan
Begin with a field dictionary and a collection of synthetic test records. Keep the calculation engine behind a separate interface from the content library. Name the settings that affect results, preserve original inputs, and document how missing information is handled. Review the rights and terms of any real software or provider before integrating it; this site includes no vendor credentials or service contract.
For an application with a backend, choose a consistent error model. Distinguish malformed input, unsupported settings, unavailable upstream data, and a valid search that returns no events. Never represent every case as a successful personal reading simply to keep the card layout full.
Version the layers independently
A calculation revision, an editorial correction, and a visual redesign are different changes. Preserve enough version metadata to explain which layer changed. If a corrected input alters the chart, invalidate dependent interpretations until they have been reviewed. If a sentence changes, do not claim the astronomical data was recalculated unless that actually happened.
Privacy starts before storage
First ask whether the feature needs personal information at all. A general reading or documentation example can work without it. A future personalized application should decide retention, deletion, access, and logging behavior before collecting records. Avoid publishing birth details in URLs or screenshots. This website has no data-entry forms and does not submit any chart information.
Useful reading paths
Start with the architecture guide, then move to natal input quality or horoscope publishing. For event workflows, compare prediction language with transit interval design. For presentation, read the accessible chart guide.