Can this existing script be improved for Siril stacking

Below is the script provided by Siril.org to process Seestar images. It was written to support Siril 1.2. Can it be improved given 1.2.4 is currently available? Thanks, Jim

############################################
#
# Script for Siril 1.2
# June 2024
# (C) Cyril Richard
# SeeStar_Preprocessing v1.3
#
########### PREPROCESSING SCRIPT ###########
#
# Script for Seestar Deep Sky images where 
# darks, flats biases are not provided. Only 
# lights are needed.
#
# Please, REMOVE all jpg files from the
# directory.
#
# If you find that too many images are discarded
# before stacking, you can increase the value after
# -filter-round= in the seqapplyreg command, line 39
#
# Needs 1 set of RAW images in the working
# directory, within 1 directory:
#   lights/
#
# Changelog
# 1.2 : using link instead of convert
# 1.3 : removing -date option: https://discuss.pixls.us/t/error-trying-to-use-seestar-preprocess-script/46400/3
############################################

requires 1.2.0

# Convert Light Frames to .fit files
cd lights
link light -out=../process
cd ../process

# Calibrate Light Frames
calibrate light -debayer

# Align lights
register pp_light -2pass
seqapplyreg pp_light -filter-round=2.5k

# Stack calibrated lights to result.fit
stack r_pp_light rej 3 3 -norm=addscale -output_norm -rgb_equal -out=result

# flip if required
mirrorx_single result

#save result using FITS keywords for the name
load result
save ../$OBJECT:%s$_$STACKCNT:%d$x$EXPTIME:%d$sec_T$CCD-TEMP:%d$degC_$DATE-OBS:dm12$
close

Given the following changelogs since 1.2.0:

siril 1.2.5
11/22/24
* Fixed crash in gaussVertical() on macOS (issue reported on pixls.us, !742)
* Fixed free space disk computation on macOS, probably for good (#1368 and #1375 and #1422)
* Fixed crash if the CWD changes after a first successfull Automated light Curve process (#1378 !746)
* Fixed bug in TIFF files that have TIFFTAG_MINSAMPLEVALUE and TIFFTAG_MAXSAMPLEVALUE set to 0 by removing these tags (#1381)
* Added HISTORY after seqcrop and GUI crop (#1142)
* Better handling of Siril opening when double-clicking on an image, under macOS (#1308, #1428 and #1140)
* Fixed potential crash in synthstar (issue reported via facebook)

siril 1.2.4
09/11/24

* Fixed CFA statistics (#1342)
* Fixed calibration command bug disabling cosmetic correction (#1348)
* Fixed calibration GUI not parsing correcting flat when using path parsing (#1348)
* Fixed non-deterministic parallel subsky RBF (#1352)
* Fixed incorrect deconvolution command arg setting (issue raised on pixls.us)
* Fixed GHT and HT not reverting previews from other dialogs (#1356)
* Fixed crash when saving image (#1353)
* Fixed execute button in wavelets not reverting preview (#1362)
* Fixed CLAHE crash on mono image (#1354)
* Fixed crash while extracting Green channel on CFA FITSEQ (#1305)
* Fixed inverted sequences on exporting to SER (#1361)
* Fixed a bug with histogram hi / mid / lo entry callback behaviour (!735)
* Fixed free space disk computation on macOS (#1368)

siril 1.2.3
06/19/24

* Fixed handling wide-char characters in filenames on Windows (cfitsio rollback) (#1324)
* Fixed compression error (files were not compressed anymore) (#1328)
* Fixed internet connection in macOS version (packaging issue)

siril 1.2.2
06/14/24

* Removed background extraction samples after using it in script (#1265)
* Fixed catalog parser problem with negative declination (less than 1°) (#1270)
* Fixed weighting by number of stars during stacking if number of stars is the same across the sequence (#1273)
* Improved mouse pan and zoom control to enable one-handed operation (!638, fixes #1271)
* Added an option to the LINK command in order to sort output by date (#1115)
* Fixed pixel size set by astrometry using binning_update preference (#1254)
* Fixed crash when querying stats on a CFA image with a selection smaller than a 2x2 square (#1286)
* Fixed crash when saving compressed and croped images (#1287)
* Disabled supernumerary use of openmp in demosaicing, which could lead to a crash (#1288)
* Fixed ser orientation error (#1258, #1261)
* Fixed crash during rejection stacking when using shift-registered sequence (#1294)
* Fixed mouse scrollwheel scrolling too fast (#1151)
* Fixed drag & drop in image display on macOS (#1310)
* Fixed bug in rgradient (Larson Sekanina) filter (#1313)
* Fixed bug in generalized hyperbolic stretches (#1314)
* Fixed path parsing error with savetif (#1318)
* Added handling of empty command pipe reads (closes #1277)

siril 1.2.1
01/26/24

* Fixed Anscombe VST noise reduction option for mono images
* Fixed HEIF import (#1198)
* Fixed Noise Reduction Anscombe VST bug with mono images (#1200)
* Fixed problems with Fourier Transform planning > Estimate (#1199)
* Fixed data initialisation bugs in copyfits() and RGB compositing tool
* Fixed exported x-column for lightcurves when Julian date is not selected (#1220)
* Fixed sampling tolerance for astrometry which was incorrectly read (#1231)
* Allowed for RA/DEC to be sorted in PSF windows (#1214)
* Added SET-TEMP as valid FITS header to be saved (#1215)
* Added configurable color for background extraction sample and standard annotations (#1230)
* Fixed argument parsing error in makepsf (!593)
* Fixed light_curve and csv export from plot when some images were unselected from the sequence (#1169)
* Added undo/redo when platesolving with astrometry.net (#1233)
* Fixed crash in findstar when detecting stars close to the border (#1237)
* Fixed using wcs info when using light_curve command (#1195)
* Allowed moving file into workdir to be picked up for livestacking (#1223)
* Fixed the way we check if there is enough space to use quick photometry (#1238)
* Fixed platesolving close to poles (#1245)
* Fixed bit depth evaluation for 8-bit images (#1244)
* Fixed division by 0 in PixelMath (#1249)

What do you want to improve in regard to 1.2.0?

Also, 1.2 means 1.2 series.

Looking to add /modify any commands that results in a better stacked image.

Yes, but looking at the changelogs, what changes/feature could justify a change in the script?

Spoil: 1.2.x releases are just bug fixes.

Thanks, I’m not an expert and was looking for the type of feedback you’ve provided.

That being said, would this change be beneficial?..

stack r_pp_light rej 3 2

You need to output rejection maps, and see how behave the rejection.

How would I do that?

https://siril.readthedocs.io/en/stable/preprocessing/stacking.html#rejection-methods