Liquid flow / bubble validity — proposed original Pico experiment

Hardware: original RP2040 Pico; official RPI_PICO MicroPython 1.26.0;
Sensirion SLF3S-1300F (3.000.091), kit-supplied 30 cm six-pin-to-pigtail cable
from EK-SLF3S-1300F (whole kit article 3.000.120). Continuity by numbered pins,
never wire colour. Single qualified 3.3 V domain: Pico physical 36 -> sensor contact 3, VDD;
Pico physical 38 -> sensor contact 4, GND; Pico GP0 physical 1 -> sensor contact 2, SDA; GP1 physical 2 -> sensor contact 5, SCL.
Two 2.2 kilohm pull-ups to that rail; 100 nF and 10 microfarad ceramic local bypass.
Sensor contact 1 IRQn and contact 6 NC individually insulated, no SCC1-USB adapter.
Water-only gravity fixture, no pump/valve/pressure test. See complete article
and slf3s-functional.svg/png before attempting a physical setup. No hardware
measurements or target binary are in this archive.

Save pico_slf3s_validity.py on Pico and explicitly run once:
  scope = {}
  exec(open('pico_slf3s_validity.py').read(), scope)
  scope['main']()
Do not call subtotal whole-run volume; no current-flow guarantee follows from
an old printed record. AIR/HIGH/SMOOTH/gaps omit intervals, then rebaseline.
STOP requests heater idle, not power-off. Clock/STOP failure can raise without
returning a report. General-call reset affects all compatible devices on the
bus: this fixture has only this sensor.

Host checks only (actual supplied program + actual vendor decoder/CRC/scales):
  python3 -m venv .venv
  .venv/bin/python -m pip install -r requirements-host.txt
  .venv/bin/python test_flow.py
Pinned Sensirion Python package 1.0.0 is a host oracle, not Pico firmware.
Tests use synthetic bus frames, no real sensor or serial port is opened.
Blank worksheet has a header only. Constructed graph data are not measurements.
