# GOAT’d Text Generator update 🐐 — context memory, new prompt guides, and a cleaner UI — retained record

> Retained reddit record from reddit:StableDiffusion.

Canonical: https://highsignal.app/data/records/9259185898f7efa3

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:StableDiffusion)  published 2026-10-27 11:10 ingested 2026-10-28 02:37

[original source ↗](https://www.reddit.com/r/StableDiffusion/comments/1xgpn4y/goatd_text_generator_update_context_memory_new/)

## retained text



```
Original post: [https://www.reddit.com/r/StableDiffusion/s/DQlE9pFWpA](https://www.reddit.com/r/StableDiffusion/s/DQlE9pFWpA)

I've added a few things to my ComfyUI text generator since my original post.

What’s new:

* Context memory: keep refining a prompt across runs without repeating everything. Includes a usage meter and a button to start fresh. Context window size by default is 8192 tokens, you can customize the value.
* Added prompt guides for Moss Image and Fibo 2.
* VGL structured JSON prompts (recommended for Fibo 2).
* Cleaner UI with separate settings and generator nodes.
* Searchable model lists by provider (api key is required).
* Expandable settings node.

GitHub: [https://github.com/ninjaGurung/ComfyUI-GOATd-Universal-Text-Generator](https://github.com/ninjaGurung/ComfyUI-GOATd-Universal-Text-Generator)

Let me know how it works for you, or if there’s something you’d like me to add!
```



document fetched 2026-10-28 00:32 · canonical https://reddit.com/r/StableDiffusion/comments/1xgpn4y/goatd_text_generator_update_context_memory_new
