easy positive

This is about as simple as a node gets: a multiline text box that outputs whatever you type as a string, labeled "positive." No encoding, no model, no magic. So why does it exist? Housekeeping. In a graph that's already a plate of spaghetti, having your prompt live in one clearly-named node - instead of buried inside a loader's widget - makes the workflow readable, and lets you route the exact same prompt text to several places at once.

Think of it as a labeled sticky note for your prompt. Its sibling easy negative does the same for the negative side. Neither one does anything clever; they exist so your prompt has an obvious home you can find at a glance and wire wherever it's needed.

How it works

There's nothing to explain about the mechanism because there barely is one: you type text, it outputs that text as a STRING. The value is what you'd expect - the parsing, encoding, and conditioning all happen elsewhere. Its whole job is to hold the text and make it connectable.

The inputs and outputs that matter

• positive - the multiline text field. Write your prompt here.
• positive (output, STRING) - the same text, on a wire.

That's the entire node. The one thing to internalize is what the output is: a string, not conditioning. You still have to send it into something that encodes it - an easy loader's positive input, a CLIP Text Encode, easy loraPromptApply, a wildcard processor. This node hands off text; the encoding is a separate step.

How to install it

Comes with ComfyUI-Easy-Use. ComfyUI Manager: search ComfyUI-Easy-Use, install, restart. Or:

cd ComfyUI/custom_nodes
git clone https://github.com/yolain/ComfyUI-Easy-Use

then install.bat (Windows) or pip install -r ComfyUI-Easy-Use/requirements.txt, and restart. No models, no dependencies of note.

Common issues & troubleshooting

"It's not doing anything to my image." Right - because it outputs a string, not CONDITIONING. If you plugged this straight into a sampler expecting conditioning, that won't connect. The output has to land in an encode step first. This is the single thing people get wrong about it.

Wrong slot on the loader. If you're feeding an easy a1111Loader / easy fullLoader, they have positive/negative prompt inputs that accept a string - wire it there, not into the sampler. Match string outputs to string inputs.

Wondering why you'd bother over just typing in the loader. You wouldn't, for a one-off. The payoff is when the same prompt (or a shared prefix) needs to reach multiple nodes, or when you want the prompt front-and-center in a workflow you'll reopen later. For a throwaway generation it's pure tidiness, and that's fine - that's all it's claiming to be.