SEO Content Layer
Clarify the use case, input requirements, and reading goal first, then continue into the tool with better context.
The API page serves people who want to connect their own LLM or agent to the site. Its job is not just to show an API key. It should explain the two working modes clearly: one mode returns raw results for your own agent to interpret, and the other uses the site analysis, spends credits, and writes the finished report back into Profile.
Searches like “divination API”, “astrology API”, “LLM chart API”, or “agent integration for readings” usually care about integration flow, credits, method choice, and report persistence, not only endpoint documentation.
These sections add more context for common high-intent questions and practical use cases.
Use raw-result mode when you already have your own LLM agent and want it to interpret the chart, cast, or spread on its own. This mode does not spend site analysis credits and does not create a site report in Profile, so it works best when Fortuneteller.Today is being used as the underlying divination engine.
Use site-analysis mode when you want the complete report generated by the site itself and also want that report to appear in Profile with the same persistence and filtering behavior as normal site usage. This mode spends credits because it runs the full site analysis workflow.
A serious agent integration is not just an endpoint list. Users also need help choosing between a concrete question, a life-trajectory reading, or a timing/topic reading. Just as importantly, birthplace, date, time, gender, question, and focus must stay exact and must not be rewritten by the agent.
The API covers every divination method available on the site. Your agent can first use the recommendation endpoint, then inspect required fields for the selected method, and finally call either the result API or the analysis workflow.
Raw-result mode does not spend site analysis credits. Site-analysis mode does spend credits because it runs the site’s own full analysis workflow.
Only API requests that use the site-analysis workflow create normal Profile reports. Those reports are automatically labeled as API and can be filtered separately.
Because that reduces bad method selection. If the agent first understands whether the user needs an event answer, a life-pattern reading, or a timing/topic reading, method choice becomes much more stable.
If you want to compare methods, question styles, or usage boundaries further, these are good next topics.