Oura (experimental): sanctioned path to display a ring-provided sleep stage timeline — without HR/RR ? #323

@ryanbr

one question in regards to the Oura Ring integration

Context

The experimental Oura integration can now pull the ring's own sleep analysis over BLE: the Gen3 ring runs its staging on-device (SleepNet) and, after wake, banks a complete night hypnogram in its event log as 0x4B/0x4E/0x5A records — 2-bit stage codes (deep/light/REM/awake).

The issue: the ring does not stream HR/RR overnight to the app (heart data arrives only while live-connected, or as banked history). NOOP's sleep pipeline — including Sleep V2 (#277) — computes sleep from continuous overnight HR/RR, so a pure-Oura night never produces a sleepSession/dailyMetric, and the Sleep card stays empty even though the ring handed us a complete, honest hypnogram.

Question

What's the way to feed an externally-staged sleep session (start/end + a stage timeline, no underlying HR/RR) into the display pipeline?

Options I can seer:

1. Reuse the stage-timeline rows path from docs: add a validation protocol for scoring changes #988 / the HC import (Android: sleep stage-timeline rows (iOS #988 port) + HC sleep import fixes #240): the Health Connect import already ingests third-party-staged sleep with stage timelines. Could a BLE source feed the same structures (as source: oura-ring / provided rather than computed)? This seems closest to "richer record wins" semantics that already arbitrate WHOOP vs imported nights. (easiest)
2. A first-class "provided sleep session" concept on the Repository — a device-supplied session that analyzeDay treats like a computed one for display, but never feeds HR-derived scoring it doesn't have inputs for.
3. Keep it out of sleepSession entirely and render a separate "ring-reported sleep" row on the Sleep card (visually distinct, clearly labeled as device-computed).

Happy to build whichever direction fits the architecture