KSampler + Config Output

This is the least exciting node in the pack and also one of the most quietly useful. KSampler + Config Output is a drop-in replacement for ComfyUI's core KSampler - same inputs, same behavior, no model modifications - except it also returns the sampler and scheduler names as plain strings alongside the LATENT. Which sounds trivial until you're trying to reproduce a song six months later and your sidecar JSON says "sampler: euler" while the actual workflow was running euler_ancestral on karras because a workflow update silently swapped the defaults.

That's the entire point of this pack, honestly: reproducibility metadata that's honest. The core KSampler gives you a latent and forgets what settings produced it. This wrapper forwards every setting to ComfyUI's common_ksampler unchanged, then hands the effective sampler_name and scheduler back out as STRING outputs you can pipe straight into MiniMax Song Metadata's sidecar.

Inputs

Identical to core KSampler: model, positive, negative, latent_image, then seed, steps, cfg, sampler_name (44 samplers), scheduler (9 schedulers), and denoise. The default cfg is 1.7, which matches the conservative guidance this pack recommends for MiniMax audio - but in the example workflow this node actually drives the FLUX.2 Klein cover-artwork branch, not the audio diffusion. If you're sampling audio with it, use whatever cfg your model family wants; the wrapper doesn't second-guess you.

One honest caveat: sampler/scheduler names are not the whole reproducibility story. The same name + seed still depends on the ComfyUI version, the backend and the device. The node reports what it was asked to do; it can't guarantee identical trajectories across machines.

Outputs

LATENT (wire it to whatever you'd wire a KSampler latent into), plus the two metadata strings. That's it. Nothing to tune, no hidden behavior.

Installing it

Standard pack install via ComfyUI Manager (search "MiniMax Music Production Toolkit") or:

cd ComfyUI/custom_nodes
git clone https://github.com/jplenio/ComfyUI-MiniMax-Music-Production-Toolkit.git
cd ComfyUI-MiniMax-Music-Production-Toolkit
python -m pip install -r requirements.txt

Dependencies: scipy, soundfile, imageio-ffmpeg, mutagen, Pillow. Restart ComfyUI and hard-refresh once.

Gotchas

Not many, because it doesn't do much. The useful habit: if you've already got a hand-built metadata system that reads KSampler widgets, this node adds nothing for you. But if you want the effective sampler/scheduler in your JSON without trusting your memory, swap your core KSampler for this one and wire the two STRING outputs into the metadata node. The one thing to know: it's KSamplerWithConfig - the old workflow name - so don't go hunting for a display name that doesn't exist.