Editing Openai/693e48dd-ce3c-800a-b636-9b3efde0fdec (section)

=== Assistant: That’s a good, controlled repeat. ===
That’s a good, controlled repeat. Here’s exactly what to watch for while you do it, so you don’t have to over-interpret anything afterward.

===== Do exactly what you said: =====
# Start scan with the phone stationary (on the dresser)
# Walk to the back door
# Pause ~10 seconds
# Walk back to the phone

Don’t chase signals, don’t rotate the phone mid-scan.

==== If anything were on you (earbud-like), you would see one line only that does this: ====
* Clear drop in RSSI as you walk away
* Flat low section while you pause
* Clear rise when you walk back
* Shape: clean U or V curve
* Magnitude: often 10–25 dB change

You already know this pattern from finding your earbud.

==== Based on your last scan, expect: ====
* Minor wobble (±3–6 dB) on fixed devices
* No clean U-shaped curve
* No single device separating itself strongly from the pack
* Lines staying clustered in the -75 to -95 dBm range

That outcome = environmental devices only.

==== If you see: ====
* jitter
* noise
* slow drift
* brief spikes that don’t line up exactly with your walking

➡️ That is not movement tracking.
Only a time-locked, symmetric curve counts.

==== When you’re back at the phone, don’t scroll immediately. Just note mentally: ====
* “Did any line clearly dip when I left and rise when I returned?”

If the answer is no (as before), that’s decisive.

Run it, then paste:
* the RSSI graph or
* the timestamps + RSSI values

I’ll help you sanity-check it, but you already know how to read this correctly.