Plain language: “A textbook Z2 ride. 92% of the time in zone, power close to flat (VI 1.01). Your heart climbed 6 beats at the same power — normal heat drift on the trainer.”

For the coach, one tap away: decoupling 1.17%, measured on minute 20–55 — warm-up and cool-down trimmed first, because an aerobic-coupling read only means something once heart rate has settled. Under 5% counts as well-coupled. If intervals.icu’s own number differs, it’s shown next to Sapere’s — never swapped in for it.

One rule-based engine writes both registers today — no AI in the loop. Every threshold it uses (a decoupling under 5%, a variability index under 1.05, an aerobic-base trend that needs at least six comparable rides) sits in code next to a test that pins it, so a wording change can never quietly move where the line is drawn. A more conversational, paid voice is designed to sit alongside this engine later — but it is only allowed to rephrase the same facts and thresholds, never to draw its own conclusion. A paying reader gets nicer language, not a different truth.

One thing it never does: attach a rule of thumb to a readiness or HRV finding. Those describe your recovery, and stay strictly descriptive — no norm, no verdict.