A native DeepSeek Harness (dsh) plugin that turns ComfyUI into an Agent-driven media generator — submit API-format or templated workflows, manage them from a side panel, and preview generated images / video in-chat.
DSH integration
Native runtime
Author-claimed
Safety audit
Unaudited
Last verified
2026-09-04
License
MIT
01What can it help you accomplish?
Let the DeepSeek Harness Agent submit and inspect ComfyUI workflows without touching the canvas
Generated images / video returned in-chat and written to the workflow asset panel
Developers running DeepSeek Harness (dsh) with a local or remote ComfyUI server
Manage ComfyUI workflows, generated assets and the task queue from the side panel
A runnable workflow library, asset preview/download, and a live queue with five-state history
dsh users who prefer a graphical panel over manually operating the ComfyUI canvas
02How to install into DeepSeek Harness
Prerequisites
- DeepSeek Harness (`web` / `desktop` profile)
- A running ComfyUI server (default `http://127.0.0.1:8188`)
Installation steps
- 01
`dsh plugin --profile web add dsh-comfyui`
$ dsh plugin --profile web add dsh-comfyui
- 02
`dsh plugin --profile desktop add dsh-comfyui`
$ dsh plugin --profile desktop add dsh-comfyui
Verify the integration
- After restart: a panel entry appears in the sidebar, a "ComfyUI" section appears in Settings, and the Agent gains all tools plus the bundled skill.
03DSH integration and capability boundaries
Native DeepSeek Harness plugin installed via the official dsh CLI; exposes ComfyUI as Agent tools plus a side panel.
Agent tools: comfyui_run / comfyui_object_info / comfyui_workflow / comfyui_skill
API-format workflows, built-in templates (txt2img / img2img / video) or workflow-library commands→Generated media, node definitions, workflow listings and skill-pack read/write
comfyui_skill can write experience back into the skill-pack directory (reused across sessions)comfyui_workflow can read/write/delete files under the local ComfyUI directory and output directorySide panel: workflows / assets / queue
ComfyUI workflows and generation results→Runnable workflow library, asset preview & download, live queue management
Load-area media loader (media proxy)
Image / video / audio assets→Content-hash-renamed load slots auto-filled into workflow load parameters
04Who is it for? When not to use it?
Good for
- Developers running DeepSeek Harness (dsh) with a local or remote ComfyUI server
- dsh users who prefer a graphical panel over manually operating the ComfyUI canvas
Not for
- Pick the matching dsh version: 0.4.0 needs DeepSeek Harness ≥ 0.1.2, 0.3.x (beta) needs 0.1.1; a mismatch makes dshmarket show a risk warning.
05Compatibility, maintenance and safety notes
- Pick the matching dsh version: 0.4.0 needs DeepSeek Harness ≥ 0.1.2, 0.3.x (beta) needs 0.1.1; a mismatch makes dshmarket show a risk warning.
- The plugin reads/writes/deletes files under the local ComfyUI directory and output directory (deleting an asset also cleans the file in ComfyUI's output directory).
- Generated files are forwarded over a same-origin route; a remote ComfyUI behind an authenticated proxy gets its key from an env var or credential store, and the key is never sent to the browser.
MIT license · last push 2026-09-01 · no published release tag (latest_release null)
06Frequently asked questions
How do I install dsh-comfyui?
Run `dsh plugin --profile web add dsh-comfyui` for the web profile, or `dsh plugin --profile desktop add dsh-comfyui` for the desktop profile. After restarting the app, a panel entry appears in the sidebar, a "ComfyUI" section appears in Settings, and the Agent gains all tools plus the bundled skill.
Which DeepSeek Harness version do I need?
0.4.0 (the `latest` tag) requires DeepSeek Harness ≥ 0.1.2 because it uses the new settings-service API; the 0.3.x beta line is for dsh 0.1.1. Pick the matching dsh version or dshmarket shows a risk warning.
What does the Agent actually do with ComfyUI?
The Agent submits API-format workflows or built-in templates (txt2img / img2img / video) via `comfyui_run`, lists supported node definitions with `comfyui_object_info`, manages the workflow library with `comfyui_workflow`, and can read/write skill packs with `comfyui_skill` so lessons learned carry across sessions.
What are the requirements?
You need DeepSeek Harness (web or desktop profile) and a running ComfyUI server (default `http://127.0.0.1:8188`). The `video` template additionally needs ComfyUI-WanVideoWrapper and the Wan 2.1 model.
Is my API key safe?
Generated files are forwarded over a same-origin route (`/comfyui/media`), so the browser never touches ComfyUI directly — no CORS, no mixed content, no API key sent to the browser. A remote ComfyUI behind an authenticated proxy gets its key from an env var or credential store (default `COMFYUI_API_KEY`).
07Related DSH workflows
modlens
by liustack
Vision bridge for text-only models: paste an image, get structured JSON evidence (OCR, layout, semantics).
agent-vision-toolkit
by anionex
Vision toolkit and skills that give text-only LLMs eyes — multi-image understanding, image Q&A, OCR, frontend UI restoration and GUI automation, with optional agent integration.
dsh-vision-router
by ysr666
Eyes for text-only DeepSeek Harness agents: built-in free vision chain (no key) + pixel-level vision tools (Q&A, grounding, crop, pixel diff, colors, OCR, SVG trace, cutout, screenshots). One-command install, no Python, image turns work like ordinary tool-calling turns.
dsh-vision-toolkit
by anionex
为纯文本 DSH Agent 提供 10 个结构化视觉工具：意图感知图片问答、长截图 OCR、原始像素 grounding、UI 还原、像素 diff 等
08Data and sources
让 DeepSeek Harness 的 Agent 智能驱动本地或远程 ComfyUI 生成任何内容。附带工作流、资产管理面板与技能包管理挂载。配套 skill 与同源媒体代理。
dsh plugin --profile web add dsh-comfyui
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-09-04. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.