Frequently asked questions

Mystilink fixes the three things that most often go wrong in metaphysics readings:

• The chart must be accurate: natal charts (BaZi, Zi Wei, astrology) are computed by a local deterministic engine, and Tarot / Liu Yao use real interactive draws, so the model never invents a chart.
• The reading must be verifiable: profile cards feed a single source of input, an uncertain birth time triggers verification before any conclusion, and complex questions are broken into a visible, traceable task list.
• Your data stays local: charts, profiles, and conversations are stored on your device (.mystilink/), with no full-library cloud sync.

In one line: it replaces the tools that invent charts, hand you template conclusions, and cannot be reproduced.

GPTs and general-purpose agents have real limits for this kind of work:

• No deterministic chart engine: they tend to invent stems, stars, or hexagrams.
• Inconsistent memory: they rarely remember your profile the same way twice.
• Hard to reproduce: there is no verification step and no traceable task flow, and conversations live in the cloud by default.

One clarification: Mystilink does not reject them — you can bring your own key (BYOK) and plug your preferred agent in as the reasoning brain. The difference is that the chart and the workflow are guaranteed by Mystilink. Don't let a general-purpose agent double as a chart calculator.

Free sites usually monetize through ads and lead generation, and that shapes what they give you:

• Template conclusions: hard to follow up on or cross-check.
• Sensitive data at risk: the birth data you enter may be retained or profiled.
• One system, split apart: charting and interpretation are separated, with no way to plug in your own model.

Mystilink differs by combining multiple systems, local deterministic charting, verification dialogue, traceable output, and data that stays on your machine by default.

Metaphysics tools are used over the long term and in private, and that shapes the pricing:

• A subscription would not fit: you would lose access and history the moment you stop paying.
• Local charting has almost no ongoing cost: charts are computed on your device, so a device license carries no real server bill.
• You pay for usage, not access: the only metered cost is LLM inference, paid through a Newton prepaid balance or your own API key (BYOK).

Separating license from usage means you pay for what you use. Note: the purchase covers the current major version (1.x); major-version upgrades may carry a separate price.