ComfyUI Node

VRGDG_PromptSplitterForFL

Pick one scene from a prompt map by index

By vrgamegirl19·Created about a year ago·Updated about 7 hours ago· 718
VRGDG_PromptSplitterForFL
    • text_output_1
    • text_output_2
    • text_output_3
    • text_output_4
    • text_output_5
    • text_output_6
    • text_output_7
    • text_output_8
    • text_output_9
    • text_output_10
    • text_output_11
    • text_output_12
    • text_output_13
    • text_output_14
    • text_output_15
    • text_output_16
    json_string{}
    index0

    VRGDG_PromptSplitterForFL is the index-based member of VRGameDevGirl's splitter family - the acronym isn't documented anywhere in the README, so treat "FL" as a workflow-internal label for whatever section this variant anchors. Unlike the fixed-count splitters that dump every scene at once, this one takes a JSON prompt map and an index, and hands you the sixteen text outputs for the scene at that index.

    That single change - an index input - makes it fundamentally different from the rest of the family. It's not "give me all scenes"; it's "give me this scene, and let my workflow loop the index." It pairs naturally with stateful steppers like the pack's VRGDG_PostRunIndexStepper, which ticks the index up after each render so the FLUX pass marches through the prompt map one scene at a time.

    How it works

    Two required inputs:

    • json_string - the prompt-map JSON as text (defaults to {}). This is the same scene-prompt structure the rest of the pack produces.
    • index - which scene to pull, 0 to 10000. Scene 0, scene 1, etc. Wire this from your index-stepper or seed node, or set it by hand to preview one scene.

    Outputs: text_output_1 through text_output_16 - sixteen STRING ports. Why sixteen for one scene? Because one scene isn't one prompt in this workflow - it's a batch of related generation parameters (positive prompt, negative, style, camera notes, whatever the FLUX section consumes), and the node unpacks the selected scene's entry across all sixteen slots. The exact field meanings aren't documented beyond the ports themselves, so treat them as "whatever the workflow section expects in that slot."

    Where it fits

    You'll see this in the part of the workflow that generates one video chunk per scene with a FLUX base image. The flow is: prompt map → subject prepender → this splitter (index-fed) → video node. Every run of the loop renders one scene, the stepper advances, and the next run pulls the next scene. If you've ever hand-copied a prompt between runs, this is the node that ends that.

    Installing it

    With the pack:

    cd ComfyUI/custom_nodes
    git clone https://github.com/vrgamegirl19/comfyui-vrgamedevgirl
    

    restart, or ComfyUI Manager → search vrgamedev. No model downloads; it's a JSON-to-text transform.

    Gotchas

    The index is zero-based, so scene 1 of your song is index: 0 - the classic off-by-one that shows up as "my first scene is the second scene." And if you wire the index stepper into it, remember the stepper only advances after a full run; previewing one scene with a manual index and then forgetting to reset it will make your video skip that scene.

    CategoryVRGDG

    Inputs (2)

    NameTypeDefaultDescription
    json_stringSTRING{}
    indexINT00–10000

    Outputs (16)

    NameTypeDescription
    text_output_1STRING
    text_output_2STRING
    text_output_3STRING
    text_output_4STRING
    text_output_5STRING
    text_output_6STRING
    text_output_7STRING
    text_output_8STRING
    text_output_9STRING
    text_output_10STRING
    text_output_11STRING
    text_output_12STRING
    text_output_13STRING
    text_output_14STRING
    text_output_15STRING
    text_output_16STRING