WIP Web UI for KV Edit over ComfyUI
Patch the Web UI to enable prompt creation, saving, drag‑and‑drop reorganization, and API integration with ComfyUI. Add support for setting the Comfy URL from the browser and plan to integrate Qwen 3.5b LLM for image description.
Patch the Web UI to enable prompt creation, saving, drag‑and‑drop reorganization, and API integration with ComfyUI. Add support for setting the Comfy URL from the browser and plan to integrate Qwen 3.5b LLM for image description.
Summary
The author has built a Web UI that interfaces with ComfyUI to provide a KV Edit experience for curated prompts. The UI allows users to create, save, and drag‑and‑drop reorder prompts, and it stores them in a cache for later reuse. The workflow includes a LoRA loader and can be exported as a ComfyUI API, enabling the UI to send prompts directly to the backend. The current version lacks the ability to set the Comfy URL from the browser, but the author plans to add this feature along with Qwen 3.5b LLM support for image description. The UI also aims to improve the forced aspect‑ratio image loading by eliminating the need for a local web server. The author encourages users to test the workflow and report bugs, and the project is hosted on GitHub with a public workflow link. The tool is still a WIP, but it demonstrates how to combine a custom front‑end with ComfyUI’s powerful API.
Key changes
- Web UI over ComfyUI for KV Edit
- Prompt creation, saving, and drag‑and‑drop reorganization
- LoRA loader integrated
- API export for ComfyUI integration
- Planned Comfy URL setting from browser
- Planned Qwen 3.5b LLM support for image description
- Improved forced aspect‑ratio image loading
- Public GitHub repository with workflow link