📜 CR Prompt List

If you've got a text file full of prompt ideas - one concept per line - and you want to run all of them without manually swapping the prompt and hitting queue over and over, CR Prompt List is the node that turns that list into a batch. It's one of Comfyroll's core list nodes and it's built specifically around a common pattern: a shared prefix, a shared suffix, and a bunch of variable middles.

What it's for

A lot of batch-generation setups follow the same shape: you've got a fixed style or quality string you want on every image ("cinematic lighting, 8k, detailed") and a fixed negative-style suffix, and the thing that actually varies between generations is the subject. Rather than retyping the fixed parts into every line of your list, CR Prompt List lets you write only the varying part once per line and have the fixed prefix and suffix applied automatically to every entry when it builds the batch.

How it works

You paste your list of subjects into the multiline box, one per line. The node splits that into rows, applies prepend_text to the front and append_text to the back of each row to build the full prompt, and outputs the whole thing as a list - meaning downstream nodes that accept a list input (like a KSampler in a batch/loop setup, or another list-processing node) will run once per row automatically, using ComfyUI's native list-execution behavior.

Inputs and outputs that matter

• multiline_text - your list of prompt bodies, one per line. This is the field you actually fill in.
• prepend_text / append_text - fixed text glued to the front/back of every row before output. Leave either blank if you don't need it.
• start_index (default 0) - skip this many rows from the top before starting. Useful for resuming a batch partway through without re-running the earlier rows.
• max_rows (default 1000) - cap how many rows get processed, in case your list is longer than you want to run in one go.

Two list outputs come out: prompt is the full assembled string (prepend + body + append) for each row, and body_text is just the raw line you typed, unmodified - handy if you want to use the plain subject elsewhere (a filename, a caption) without the style boilerplate attached. Plus the standard show_help link to the node's wiki page.

How to install it

Install via ComfyUI Manager (search "Comfyroll Studio"), or clone it directly:

cd ComfyUI/custom_nodes
git clone https://github.com/Suzie1/ComfyUI_Comfyroll_CustomNodes.git

Restart ComfyUI afterward. No models or heavy dependencies - this is pure text handling.

Common issues & troubleshooting

Only one image generates instead of a batch. This almost always means a downstream node isn't actually consuming the list - check that whatever you connected prompt or body_text into is set up to handle list inputs (in native ComfyUI, connecting a list output into most sampler-adjacent nodes triggers automatic per-item execution, but some custom nodes expect a single value and will only grab the first item).

Prepend/append text showing up doubled or in the wrong place. Remember these apply to every row identically - if your multiline text already includes the style boilerplate on each line, you'll get it twice. Keep the fixed parts out of the multiline box and only in prepend_text/append_text.

Batch runs way more (or fewer) generations than expected. Check start_index and max_rows - it's easy to leave max_rows at its default of 1000 and accidentally queue far more than intended if your text box has more lines than you remember, or to have start_index set from a previous test run and silently skip rows.

Pack doesn't load at all. If no Comfyroll nodes show up in the node menu, that's typically an install problem rather than anything specific to this node - reinstall the pack cleanly (delete the folder, re-clone, restart) rather than debugging around a partial install.