NO8D-Prompt-libraries (Patreon)

Required Nodes

https://github.com/no8d/ComfyUI-NO8D-controls

Member Exclusive Prompt Library

https://www.patreon.com/collection/2293418

As the title suggests, this is an upgraded version of the NO8D-Krea2 Style Selector node. If you haven't used it before, just think of it as a loader for a customizable prompt library. The reason I consider it important is that I've noticed many people struggle with writing prompts: every time a new major image generation model is released, there's a wave of repeated prompt testing, as people are eager to find out what kind of prompts can unlock the new model's full potential, yet few actually collect these prompts systematically. As a result, every time they want to write something, they end up having to ask GPT for help — that's why I created this node.

When you first call it, you might feel a bit confused, as it's completely empty — that's exactly why I wrote this article, to help everyone get the hang of it easily. First, you need to click that blue Settings button.

You will see a pop-up window, which serves as the foundation for all your future lexicons. I will add more lexicons later. Now let's take a look at the built-in lexicons I have prepared for you — four sets of style prompt lexicons, all embedded within the node installation package. Click "Import Lexicon" to load the lexicon and follow the file path to complete loading.

Of course, you can also load multiple lexicons, and you can view the name and entry count of each lexicon.

If you open these xlsx files with work tools, you will find that they are pre-edited prompt libraries, where you can clearly view the schematic diagram, name and content of each prompt.

Let's go back to ComfyUI, and now you can see all the lexicons you've loaded in the node. At this point, you've finished the hardest part of this node. lol.

The drop-down list contains all the lexicons you just loaded. The functions of the "favorites" and "history" buttons next to it are exactly as their names suggest (duh, lol). It's worth noting that clicking the button once activates the function, and clicking it again turns it off. The window below displays the lexicon content by default.

The second row features the search function, which supports searching for prompt words in the currently loaded lexicon, while the two function buttons next to it are "Output All" and "Output Randomly" respectively. We will explain the reason for these two buttons later.

Finally, there is the preview window, where you can view all the flashcards in the currently loaded word bank.

It supports the operation of shortcut keys combined with the mouse:

Arrow keys: ↑↓←→

翻页：page-up/page-down

Home/End: home/end

Multiple selection: Ctrl + left mouse click

Multiple selection: Shift + Left mouse button

It also supports right-click operations, with different right-click functions applicable to different windows, and further variations when right-clicking on word cards versus blank areas.

Double-click a single flashcard in the preview window to open its edit page, where you can modify the prompt word, title and schematic diagram, or save it as a new file to create your own flashcard.

If you have created a large number of your own flashcards, I recommend exporting them in the settings, so that you will have a custom word bank. Of course, I guess you will put all of them in the favorites folder, so I have also set up the export function for the favorites to make it convenient for you to save them.

Speaking of favorites, do you still remember the two function buttons next to the search box? They work on the current preview window. If you open the favorites and activate "Output All / Output Random", you will send a list of multiple prompts to the downstream. The benefits of doing this go without saying. Oh, by the way, if any flashcard is selected, only the prompt of that flashcard will be output. To achieve output all / output random, you need to click the flashcard again to deselect it, so please pay attention to this point.

Finally, if you have a large number of lexicons, you can build a prompt structure node group. To this end, I have specifically developed a prompt merging node, and the combination of them is roughly as follows.

1: A simple description, here I'm using my icon

2: I have loaded a set of emoji lexicon

3: Style lexicon built into NO8D-control

4: Connect them to the prompt merging node

5: You will clearly see a three-part prompt, which is arranged in the same order as they were input.

Click to generate, and then you'll get an image like this. ..

Looking at this image, I guess I won't be sleeping too well tonight. .. hahaha. .. It's okay, I've already put the sample workflow into the node package, and you can get it by upgrading to the latest version of the node package.

OK! That's all. I hope you'll like it.

🫡 May the AI-power be with you, see you soon !🫡

Disclaimer: My model or workflow is solely for learning and communication purposes and shall not be used for any commercial activities. All responsibilities for the generated content, including its legality, ethical implications, or potential misuse, are borne entirely by the user. By using this model or workflow, you agree to this principle. Please use it responsibly within the scope of local laws.