Ran a week of prints straight from the terminal, not going back

I tried something similar on my Centaur setup last winter when I was running a huge batch of enclosures for a project. I hated the idea of sitting in Cura just adjusting retraction settings for every single file. The constant manual input was a real drag when you're trying to throughput a lot of prints. I ended up writing a simple Python script that reads the model directory and then pipes the required parameters to the slicer CLI for each one. It saved me at least an hour of clicking through the settings menu. My biggest unexpected hurdle was getting the pathing right when the slicer accessed the local print profile files. It took a couple of tries to get the relative paths in the script to resolve correctly.

I found that automating the start and end g-code for temperture calibration steps worked incredibly well, though. Instead of manually initiating the heat-up sequence and waiting for the thermal stability check, the script handled the timing and logged the results directly into a spreadsheet. It's definitly more satisfying to see the whole thing execute with just one command than clicking through twenty diferent windows. It makes the whole process feel much more systematic.

I've been running a workflow that’s more manual than what you described, but it works well for the complex jobs I usally print on my Voron. I am not using any wrapper around the slicer, which is obviously slower, but I find it gives me much more granular control over the final G-code. My biggest issue isn't the slicing process itself, but how the initial mesh and support generation are handled for large, open-frame projects. For instance, when I print those big internal cooling ducts for my setup, the standard slicers often generate internal supports that are too chunky, which messes up the surface finish.

So, I have a custom pre-processing script that takes the STL and uses MeshMix to slightly alter the model geometry before it ever hits the slicer. It’s about removing unnecessary overhang points and slightly smoothing sharp corners that tend to cause ugly scarring when the print moves. I usually run this through a simple command prompt sequence and then load the result into PrusaSlicer for the actual print profiles. It’s a two-step process instead of one, but the final results are way cleaner. I found that the reduced surface texture on the model input dramatically lowered the required infill density for those specific components. This wasn't about automating the process into a single line, but about optimizing the input model quality to make the entire print cleaner and faster on the printer itself.

The plain text angle is the real sell for me. I keep my slicer profiles in a git repo now, so when I tweak something on my 2.4 I can diff it against what worked six months ago and roll back if a print comes out ugly. Last month I traced a stringing issue back to a profile change I'd made three weeks earlier and completely forgot about. Took two minutes to find instead of an afternoon of guessing. Curious how it handles per-object settings though, does it expose that or is it profiles only?

I'm still pretty new to all this and half the time I don't even know what setting I'm supposed to change, I just poke at the little boxes in the app and hope for the best. The bit about it reading the bed size and nozzle off the printer on its own sounds like magic to me. Does that work on any printer or does it only do that on certain ones? My machine is a cheap hand-me-down and I had to measure the plate myself with a ruler when I set it up.