AI Agent Hub
Back to plugins
🧰

dsh-comfyui-ctl

Web Tools Updated 2026.09.01

Run the following command in DeepSeek Harness:

dsh plugin install liyi3068238601-oss/dsh-comfyui-ctl

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install liyi3068238601-oss/dsh-comfyui-ctl in the DeepSeek Harness terminal to install; full source at https://github.com/liyi3068238601-oss/dsh-comfyui-ctl

About this plugin

You are deep into a ComfyUI session and keep alt-tabbing to the browser to check the queue, cancel a job, or hunt down last night's output folder. dsh-comfyui-ctl folds that whole loop into a single chat turn: ask your AI partner what is running, kill the third item, pull up the most recent 768x1152 render. It reads the ComfyUI HTTP API directly and hands you the answer without leaving the conversation.

The plugin exposes a dozen-plus tools spanning queue inspection and cleanup, task interruption, history and output-image retrieval, model inventory (checkpoints, LoRAs, VAEs, CLIPs), node and extension listing, system resource stats, local image upload, and full API-format workflow submission with optional synchronous wait. The core transport logic lives in a zero-dependency ComfyClient module that is not coupled to the DSH host, making it straightforward to wrap in an MCP server or reuse in other automation stacks.

It is a natural fit for creators who run ComfyUI locally and want natural-language control over their entire generation pipeline, and for developers who need to wire a self-hosted ComfyUI instance into an AI workflow. By default the plugin targets 127.0.0.1 only, and local-path uploads require an explicit whitelist, keeping the security posture tight for trusted local environments.

Use Cases

  • Query ComfyUI queue status and interrupt or clear tasks via natural language
  • Submit API-format workflows conversationally and wait for results in-line
  • Inventory local model folders, upload reference images, and retrieve output renders

Best For

  • AI image creators running ComfyUI on local hardware
  • Full-stack developers wiring a self-hosted ComfyUI into automation pipelines
  • Power users who prefer conversational control over constant browser tab-switching