Pre-registered protocol: Felt Intelligence reads scored by the person. Hypotheses H4 to H7, count C1, frozen lexicon and scoring rules. No results reported.

This is a pre-registration, not a result. It records, before any scoring, the tests a small research project will run on Felt Intelligence (FI), a reading of a person that the person scores: a dated line about their own past is spoken, they answer "That's me" or "Not quite", and their answer outranks the reading. Four hypothesis tests (H4, H5, H6, H7) and one descriptive count (C1) are stated with their populations, windows, measures, pass rules, readability thresholds and fail actions. H4 to H6 were first published on satya.org on 17 and 18 October 2026 with SHA-256 fingerprints; H7 and C1 were written on 5 November 2026. The identical text was registered on the Open Science Framework on 5 November 2026 at 10:41 UTC (osf.io/krs8a). H7 is unrun; its window opens only when the shuffled-delivery build is live, and that date will be published before any tap counts. The window label on the first-reflection count (529) is being settled and logged before a single claim is scored. No outcome is promised. Nothing here is called proven. Reads are published pass, fail, not yet decisive, or "not enough people yet", with denominators, at satya.org/research.