BatchCreativeInterpolationNodeDynamicSettings

BatchCreativeInterpolationNodeDynamicSettings is the pack's most mysterious node, so let's be honest about what it is up front: it's a settings generator for some other pack's batch-interpolation machinery, not a sampler and not an image node. It takes a bunch of per-frame tuning parameters, accumulates them, and spits out three comma-separated STRING schedules. You will almost certainly never add this node to a graph deliberately - you'll meet it because you loaded a workflow that uses it. Zero search impressions, nothing in the README about it, no community signal. It's an artifact of the pack author's own experimentation.

What it does mechanically. The image_count input (1–50) selects how many "keyframes" you're configuring. For each of the first 49 slots the node exposes four parameters:

• frame_distribution_N (4–64) - how many frames this keyframe spans. These are accumulated: the output for slot N is the running total of distributions up to N, not each value in isolation.
• key_frame_influence_N (0–10) - how strongly that keyframe steers the interpolation.
• min_strength_value_N / max_strength_value_N (0–1) - the strength range for that keyframe.

Outputs are three STRINGs: dynamic_frame_distribution_values, dynamic_key_frame_influence_values, and dynamic_strength_values. The distribution output is the cumulative frame positions; the influence output is the per-keyframe values; the strength output is pairs in (min,max) form, e.g. (0.0,1.0),(0.0,1.0). Comma-separated, trailing commas trimmed - formatted to be pasted or wired into whatever interpolation node consumes these schedules.

How you'd use it. Only one path makes sense: you have the interpolation nodes this feeds (likely another custom pack's "batch creative interpolation" node), and you want a compact UI for defining per-keyframe timing and strength without hand-typing schedule strings. Set image_count to the number of keyframes, tune the per-slot values, wire the three STRING outputs to the interpolation node's schedule inputs. If the target node expects a different format, you'll find out immediately - this pack provides no consumer for these outputs, so the contract is whatever the other pack documents.

The honest verdict. The defaults (distribution 4 for slot 1, 16 everywhere else, influence 1, strength 0–1) suggest the author tuned it for some specific interpolation pipeline, and that's about all we can infer. It's not documented, it's not discussed, and there's no guarantee it matches the interpolation pack you're using. Treat it as a "if the workflow author included it, leave it alone and just tweak the numbers" node - and if you're building an interpolation workflow from scratch, look at the interpolation pack's own scheduling nodes first, because they'll document their input format.

Install. It rides along with the pack:

ComfyUI Manager → search "komojini-comfyui-nodes" → Install → restart

or cd ComfyUI/custom_nodes && git clone https://github.com/komojini/komojini-comfyui-nodes. No models. If you loaded a workflow and hit this as a "missing node," installing the komojini pack is the fix - that's the most common way beginners end up here anyway.