Action List (kppb)
What the subject is doing — as toggles, not a sentence
- list
- count
There's a meaningful difference between a pose and an action, and this node is on the action side. A pose is how the body is arranged - standing, sitting, arms crossed. An action is what the subject is doing: drinking coffee, taking selfie, applying lipstick, checking phone, typing on laptop, blowing a kiss, holding shopping bags, reading book, listening to music with earbuds, writing in journal, winking. Action List (kppb) is the checkbox panel for the "doing" verbs - 32 of them - built for batch exploration.
Why it exists separately
The Prompt Builder has both a pose dropdown and a free-text action field, and they compose differently: pose is static framing, action is narrative. This list feeds the narrative side. Same character, same scene, same lighting, rendered while drinking coffee versus while laughing at her phone - that's a very different photo. Ticking five actions and queueing them is the fastest way to find which slice of life your character actually lives in.
How it works
Identical to the other list nodes in this pack: every action is a boolean toggle, custom_entries lets you add your own verbs (one per line, e.g. "feeding a pigeon"), and the outputs are:
list- the selected action strings.count- how many, for sizing the batch.
Wire list into a ComfyLab XY Plot Queue dimension and count to its step count. Since these strings are short verb phrases, they also read fine if you just append them to the Prompt Builder's action field manually for a single shot.
Installing
Pack standard:
cd ComfyUI/custom_nodes
git clone https://github.com/artokun/ComfyUI-Photoreal-Prompt-Builder.git comfyui-klein-photoreal-promptbuilder
Restart, no pip deps, node under conditioning/klein - or search "Photoreal Prompt Builder" in ComfyUI Manager.
Two gotchas worth naming. The ComfyLab one, as always: install "ComfyLab Pack" manually because Manager can't auto-resolve the XY Plot queue nodes. And the semantic one: several actions overlap with poses (taking selfie vs the mirror-selfie poses), so if your grid comes out weirdly similar, check you're not stacking an action and a pose that describe the same thing - pick one axis for it.
Inputs (33)
| Name | Type | Default | Description |
|---|---|---|---|
| adjusting earring | BOOLEAN | false | — |
| adjusting sunglasses | BOOLEAN | false | — |
| applying lipstick | BOOLEAN | false | — |
| blowing a kiss | BOOLEAN | false | — |
| brushing hair | BOOLEAN | false | — |
| checking phone | BOOLEAN | false | — |
| drinking coffee | BOOLEAN | false | — |
| drinking wine | BOOLEAN | false | — |
| eating | BOOLEAN | false | — |
| fixing hair in mirror | BOOLEAN | false | — |
| holding bouquet | BOOLEAN | false | — |
| holding shopping bags | BOOLEAN | false | — |
| laughing at phone | BOOLEAN | false | — |
| licking lips | BOOLEAN | false | — |
| listening to music with earbuds | BOOLEAN | false | — |
| looking in mirror | BOOLEAN | false | — |
| opening gift box | BOOLEAN | false | — |
| playing with necklace | BOOLEAN | false | — |
| pouring drink | BOOLEAN | false | — |
| putting on jacket | BOOLEAN | false | — |
| putting on shoes | BOOLEAN | false | — |
| reading book | BOOLEAN | false | — |
| removing jacket | BOOLEAN | false | — |
| scrolling phone on couch | BOOLEAN | false | — |
| sipping through straw | BOOLEAN | false | — |
| stretching after waking | BOOLEAN | false | — |
| taking selfie | BOOLEAN | false | — |
| tying shoelaces | BOOLEAN | false | — |
| typing on laptop | BOOLEAN | false | — |
| winking | BOOLEAN | false | — |
| writing in journal | BOOLEAN | false | — |
| zipping up dress | BOOLEAN | false | — |
| custom_entries | STRING | — |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| list | LIST | — |
| count | INT | — |