← sources

comfyui Agent

source: reddit (reddit:StableDiffusion)published 2026-09-19 15:05ingested 2026-09-20 02:37

retained text

**Has anyone built an autonomous AI agent for local AI workflows?**

I’m wondering if something like this already exists.

For example, I give the agent a script and tell it to split it into a specific number of lines/scenes, then automatically figure out which local image/video models I have available and choose the right ones.

If a model is slow, I’d want the agent to research whether there’s a way to speed it up, check the available options, modify the workflow, test it, and improve it automatically.

Basically, I want to give the agent a high-level task like **“turn this script into a video”**, and have it handle the whole workflow by itself — choosing models, building the workflow, generating images/videos, troubleshooting errors, modifying nodes/settings, and refining the workflow when needed.

Has anyone built something like this using ComfyUI, local LLMs, or other AI agent frameworks? 

document fetched 2026-09-20 00:30 · canonical https://reddit.com/r/StableDiffusion/comments/1wkolo6/comfyui_agent