Nodes/Powder Nodes/Powder Prompt List
ComfyUI Node

Powder Prompt List

Batch several scenes without duplicating the graph

By E2GO·Created 7 months ago·Updated 4 months ago· 0
Powder Prompt List
    • positive_prompts
    • negative_prompts
    prompt_config[]
    prompt_1_text
    prompt_2_text
    prompt_3_text
    prompt_4_text
    prompt_5_text
    prompt_6_text
    prompt_7_text
    prompt_8_text
    prompt_9_text
    prompt_10_text
    prompt_11_text
    prompt_12_text
    prompt_13_text
    prompt_14_text
    prompt_15_text
    prompt_16_text
    prompt_17_text
    prompt_18_text
    prompt_19_text
    prompt_20_text
    negative_1_text
    negative_2_text
    negative_3_text
    negative_4_text
    negative_5_text
    negative_6_text
    negative_7_text
    negative_8_text
    negative_9_text
    negative_10_text
    negative_11_text
    negative_12_text
    negative_13_text
    negative_14_text
    negative_15_text
    negative_16_text
    negative_17_text
    negative_18_text
    negative_19_text
    negative_20_text

    The Powder Prompt List is the boring workhorse of the E2GO "Powder" pack, and that's a compliment. Its whole job: hold up to 20 prompt+negative pairs on one node, let you enable or disable each one, and output them as two lists. It's the "generate five different scenes with the same LoRAs and styles" node - the thing you wire up when you're tired of swapping text into a single CLIPTextEncode between runs.

    Everything in ComfyUI that takes a list of prompts gets awkward fast with stock nodes. You end up with a row of text primitives, a bunch of concatenation, or a pile of duplicated sampler branches. This node collapses that into one compact node whose output shape matches what Powder Lora Loader and Powder Conditioner expect natively.

    How it works

    There are no wire inputs on this node - only the prompt_config slot (a JSON string the frontend manages for you) and the up to 20 prompt_N_text / negative_N_text fields that the + Add Prompt / - Remove Prompt buttons create. Each slot has an Enabled toggle and a multiline text field for the prompt and the negative.

    On execution, it collects every enabled slot and returns the positives and negatives as two STRING lists. The lists are aligned - slot 2's negative pairs with slot 2's positive - and the node chain after it treats the lists as one unit.

    The output feeds straight into the Powder pipeline: the Lora Loader consumes the prompt list and multiplies it against your LoRA combos, and the Conditioner encodes the results with caching, so re-running with the same prompts doesn't re-encode.

    The inputs that matter

    Honestly, there's one: Enabled. The most common workflow is to write a bunch of scene prompts, then toggle slots on and off to test subsets without deleting text. Since the node only outputs enabled slots, you can keep a growing library of scene prompts in the graph and generate whatever subset you care about today.

    Everything else is just the text fields - and note that negatives are per-slot, so you can give each scene its own negative rather than one global string. That's genuinely nicer than the wildcard node for scenes that need different exclusions.

    Outputs

    • positive_prompts - STRING list, one per enabled slot.
    • negative_prompts - STRING list, aligned with the positives.

    Installation

    One install for the whole pack:

    cd ComfyUI/custom_nodes
    git clone https://github.com/E2GO/e2go-comfyui-nodes.git e2go_nodes
    

    Restart ComfyUI, or install via ComfyUI Manager by searching e2go_nodes / E2GO. No extra pip dependencies; targets ComfyUI 0.17+.

    Common issues

    The main trap is expecting a disabled slot to output an empty string. It doesn't - it outputs nothing, and the lists re-align. If you disable slot 2 of 3 and the third prompt suddenly pairs with the wrong negative, that's why: the lists only contain enabled slots. Keep disabled slots at the end of your list, or get comfortable with the re-alignment.

    Second, this node has no text-processing smarts. No wildcard expansion, no line-splitting, no prefix/suffix logic - if you want to paste a big block of text where one line = one prompt, use Powder Prompt Wildcard instead. They output the same shape (positive/negative STRING lists), so they're drop-in interchangeable in a workflow; the usage guide calls that out explicitly. The Prompt List is for curated, per-slot, enabled/disabled scene management; the wildcard is for bulk text. Choose based on how you think about your prompts.

    Categorye2go_nodes

    Inputs (41)

    NameTypeDefaultDescription
    prompt_configSTRING[]
    prompt_1_textoptSTRING
    prompt_2_textoptSTRING
    prompt_3_textoptSTRING
    prompt_4_textoptSTRING
    prompt_5_textoptSTRING
    prompt_6_textoptSTRING
    prompt_7_textoptSTRING
    prompt_8_textoptSTRING
    prompt_9_textoptSTRING
    prompt_10_textoptSTRING
    prompt_11_textoptSTRING
    prompt_12_textoptSTRING
    prompt_13_textoptSTRING
    prompt_14_textoptSTRING
    prompt_15_textoptSTRING
    prompt_16_textoptSTRING
    prompt_17_textoptSTRING
    prompt_18_textoptSTRING
    prompt_19_textoptSTRING
    prompt_20_textoptSTRING
    negative_1_textoptSTRING
    negative_2_textoptSTRING
    negative_3_textoptSTRING
    negative_4_textoptSTRING
    negative_5_textoptSTRING
    negative_6_textoptSTRING
    negative_7_textoptSTRING
    negative_8_textoptSTRING
    negative_9_textoptSTRING
    negative_10_textoptSTRING
    negative_11_textoptSTRING
    negative_12_textoptSTRING
    negative_13_textoptSTRING
    negative_14_textoptSTRING
    negative_15_textoptSTRING
    negative_16_textoptSTRING
    negative_17_textoptSTRING
    negative_18_textoptSTRING
    negative_19_textoptSTRING
    negative_20_textoptSTRING

    Outputs (2)

    NameTypeDescription
    positive_promptsSTRING
    negative_promptsSTRING