add prompt 2 to being error propagation

Excellent. Would you consider 6.4 and explain why the scan server does not deliver the complete waveform of Data:XPROFILE or Data:XY to the FittingIOC? Also, would you consider 6.6 and suggest an edit to the calls to matplotlib such that the images are rendered in the 640x480 dimensions?

Claude updated the document.

## Prompt 2

You are working in the tasking project, bl4b-alignment-integration branch. The code uses the Experimental Physics Industrial Control System (EPICS) as its basis. Please read tasking/analysis-chis-scan-integration.md to orient yourself. I received instructions on uncertainty propagation from a collaborator. Please read PoissonStatistics.mhtml. The task for this session is to modify the code in /home/controls/common/scantools/issue/4850/ involved in the chis scan to *correctly* compute the ErrSigmaX and *correctly* propagate this error thru the operations that those points undergo. Please plan the edits as I am interested in the process so that I can learn the steps. As I understand it, the uncertainty will propagte according to the standard deviation formula for *each* chis scan point, correct? As always, if you need a tool to analyze the code or artifacts so far, please install it yourself according to the instructions in the cross-project layer (use pixi or uv with venv). If you are missing a log file or piece of information, please ask me to fetch it for you. If you run into an issue where you need sudo, ask me to execute the line in a separate terminal.