🌻 CSV Reader

Reads a CSV file and lets you pull one row per generation. The classic use is a spreadsheet of prompts: put one prompt per line, and this node hands you the next one each run so you can batch through a whole list without touching the graph. It's the backbone of any "generate 200 images from my prompt sheet" setup.

Why a CSV instead of a text box

Because a spreadsheet is a far nicer place to manage a big list than a cramped node widget. You can sort it, edit it in Excel or Google Sheets, keep columns for prompt / negative / style, and version it as a file. The CSV Reader is the bridge between that spreadsheet and your workflow - it reads the file and picks the line you want.

The column_index input is what makes the multi-column part work: keep prompts in column 0, negatives in column 1, and pull whichever you need by index.

The inputs

• file_path - path to the CSV (on the machine running ComfyUI).
• column_index - which column to read (0 is the first).
• seed and increment - together these advance which line you get. increment sets the step; the seed carries the position, so each run walks forward through the file.
• index (optional) - jump straight to a specific line when you want a particular row rather than the automatic progression.

The outputs

Seven, but you'll mostly care about a couple:

• Selected_Line - the single row you landed on. This is the one you wire into your prompt encoder. It's the whole point of the node.
• Lines - every row as a list, if you'd rather iterate the entire file at once.
• Total_Lines - the row count, useful for setting up a loop or knowing when you've reached the end.
• Full_Content - the raw file as one string.
• File_Path, Seed, Increment - pass-throughs, handy for chaining the progression onward.

Using it for a prompt sweep

Set file_path to your sheet, column_index to your prompt column, wire Selected_Line into CLIP Text Encode, and queue a batch - each run steps to the next prompt. If a run lands on a blank or malformed row you'll get odd output, so keep the CSV clean: no stray empty lines, and watch commas inside a prompt, since a naive CSV split treats a comma as a column break. Quote fields that contain commas, or keep prompts in their own dedicated column. One quick test run against Total_Lines confirms it's reading the file the way you think.

Installing ComfyUI-Addoor

ComfyUI Manager: Install Custom Nodes → search "ComfyUI-Addoor" → install → restart. Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/Eagle-CN/ComfyUI-Addoor.git
cd ComfyUI-Addoor
pip install -r requirements.txt

Restart and find it under 🌻 Addoor / CSV. Pair it with the Incrementer if you want tighter control over which line comes next.