How we verify
A reading you could reject.
Every astrology app survives on the same trick: tell you something so vague that no version of you could disagree with it. Self: Prime is built so you could.
“You have a deep inner strength.” “You value authenticity.” True of you, true of your neighbor — recognition of everyone is recognition of no one. A sentence you can’t disagree with carries no information about you.
Your reading makes claims about how you specifically are built — which means every one of them could be wrong. That isn’t a risk we tolerate; it’s the design. A line you could reject is the only kind worth reading. Here is the machinery that holds us to it.
01 · The engine
The same chart, every time.
Your chart is calculated, not curated. Your birth date, time, and place run through a deterministic engine — astronomy and arithmetic, no randomness, no editorial hand. The same three facts produce the same chart today, tomorrow, and on any device, down to the line.
The artifact
Two reference charts — one born August 1979 in Tampa, one born September 1983 in Naples — are asserted in our automated test suite. Every release recomputes both from raw birth data and checks the result against the published reference, placement by placement. If a code change moves a single gate on either chart, the release fails.
02 · The rule
Every line must be one you could reject.
Before a word of interpretation is written, the writing model is bound by a standing rule — the same one, on every reading:
Every interpretive line must be falsifiable: a differently-built person — a different type, a different authority, the opposite centers defined — would read it and reject it as untrue of them. If nobody could reject the line, it is filler — cut it, or sharpen it until someone would.
The rule comes with a blacklist. These are banned, by name:
- “You have a deep inner strength.”
- “You value relationships.”
- “You have untapped potential.”
- “At your core, you seek balance.”
The artifact
Every claim in your reading must trace to a named mechanic of your chart — your type, your authority, a specific center or channel — so you can see exactly why it was said about you, and reject it if it misses.
03 · The judge
A second model cross-examines the prose.
Before your reading is shown, a second, independent model is handed the finished prose — and nothing else. No chart. No birth data. Its job is to reconstruct you from the words alone: your type, your decision-making authority, which of your nine centers are open. We then compare its reconstruction to your actual chart. The match is your grounding score.
Vague writing fails this exam by construction — you cannot rebuild a specific chart from sentences that would fit every chart.
On the deep reading, the verdict is enforced: if the blind reconstruction misses, the prose goes back to be repaired against your real chart before it ships. On the standard reading, the score is measured and saved with your reading — and when it clears the bar, it’s printed right on it:
Grounding 0.83 — independently verified
The artifact
That chip, on your reading, is the judge’s scorecard: how well an independent model reconstructed your chart from the prose alone, on a 0–1 scale. It only appears when the match clears 0.70 — we don’t decorate the ones that don’t.
See what survives cross-examination.
Show me my chart →Free. Three fields. No account.