Note that putting R3/R4 voltage divider across a 3V power rail (2x AA cells) results in a constant current drain of 1.2uA in addition to the MCU STOP current of around 2uA.
If you’re using the same 3V Vbatt for the MCU, there’s no problem with exceeding the rail, so you might consider removing the divider and measuring the voltage directly.
If you’re using Vdd as the ADC reference (as the MCU does) then you’re in a predicament, right? You’ll always read full-scale on the ADC because the reference is the same as the value you’re measuring. Never fear, ST thought ahead for this: there’s an internal calibrated reference in the MCU around 1.2V IIRC. So you read the calibrated reference and do some math (explained in the STM32WLxx documentation) to figure out what the Vdd rail is.
This means you don’t need any pin to measure Vbatt, you just measure the internal reference and calculate Vdd from that. Note that this calibration is required anyway when reading external voltages. This also eliminates inaccuracy due to component tolerance of R3/R4.
I don’t know if RUI does that or supports doing this, but it’s easy to do with STM32Cube.
Warning: I just thought this up before having the first cup of coffee for the day. I might have missed a detail (but pretty sure I didn’t).