Apache-2.0 native DeepSeek Harness image plugin: agent-driven text-to-image / image-to-image generation, iterative `/edit_image` editing, multi-model compare, templates and gallery, plus an e-commerce product-visual mode — installed via `dsh plugin --profile web add`.
DSH integration
Native runtime
Author-claimed
Safety audit
Unaudited
Last verified
2026-09-03
License
Apache-2.0
01What can it help you accomplish?
Generate images from a prompt inside DeepSeek Harness, then keep editing them in the same workflow
Generated images viewable in the workbench, with `/edit_image` iterative editing and shared cross-device history
DeepSeek Harness (dsh web) users who want agent-driven text-to-image and image-to-image generation without leaving the chat
Build a full set of e-commerce product visuals from a single product photo
Main image, selling-point image, scene image and more, generated as a complete product visual set
E-commerce sellers and marketers using DeepSeek Harness who need batch product imagery
02How to install into DeepSeek Harness
Prerequisites
- DeepSeek Harness installed
- Node.js 20+
Installation steps
- 01
Run `dsh plugin --profile web add @dickpy/dsh-imagegen`
$ dsh plugin --profile web add @dickpy/dsh-imagegen
- 02
Restart `dsh web` after install
$ dsh web
- 03
Open Settings → Plugins → AI image generation, add a provider (API URL + key), click Detect available models, pick the image model and save
- 04
Open the Generate tab, type a prompt and generate your first image
Verify the integration
- Open the Generate tab, type a prompt and successfully generate an image — that confirms the plugin and provider are configured correctly
03DSH integration and capability boundaries
Installed via `dsh plugin --profile web add`, runs as a native DeepSeek Harness (dsh web) AI image workbench
Agent chat generation & /edit_image editing
prompt, conversational instruction, or reference image→generated image, iteratively editable in the same workflow via /edit_image
Side-by-side multi-model compare
same prompt across multiple image models→side-by-side results from multiple generation models
E-commerce product visual mode
one product photo→main image, selling-point image, scene image and more, as a complete product visual set
Host-process queued generation
image generation tasks→queued generation results that never block the chat
generation tasks are queued by the host process (host-process queue, UI stays responsive, conversation is not interrupted)
04Who is it for? When not to use it?
Good for
- DeepSeek Harness (dsh web) users who want agent-driven text-to-image and image-to-image generation without leaving the chat
- E-commerce sellers and marketers using DeepSeek Harness who need batch product imagery
Not for
- Requires a configured OpenAI-compatible image endpoint (API URL + key); without it no images can be generated. Also requires an existing DeepSeek Harness install and Node.js 20+.
05Compatibility, maintenance and safety notes
- On Windows, if PowerShell's script policy blocks the `dsh plugin` command, run `dsh.cmd` instead.
- Requires a configured OpenAI-compatible image endpoint (API URL + key); without it no images can be generated. Also requires an existing DeepSeek Harness install and Node.js 20+.
Apache-2.0 · maintained (latest release v1.5.2, 2026-09-02)
06Frequently asked questions
How do I install dsh-imagegen into DeepSeek Harness?
Run `dsh plugin --profile web add @dickpy/dsh-imagegen`, then restart `dsh web`. The sidebar's 'New chat' entry becomes a dual 'New chat / Generate' tab. On Windows, if PowerShell's script policy blocks the command, use `dsh.cmd` instead.
Can the DeepSeek Harness agent itself draw pictures, not just me?
Yes. After install, keep 'Allow agent to call image generation' enabled (it's on by default). The agent can then generate images during a conversation, and you can continue editing any result with the `/edit_image` slash command.
What image provider does it need?
Open any OpenAI-compatible image endpoint: open Settings → Plugins → AI image generation, add a provider with its API URL and key, click 'Detect available models', pick the image model, and save. Without a configured provider no images can be generated.
Does it interrupt my DeepSeek Harness chat while generating?
No. Generation tasks are queued in the host process, so the UI stays responsive and the conversation is never blocked, and you don't have to move images between separate tools.
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
`dsh-imagegen` 是 DeepSeek Harness（DSH）的原生 AI 图像工作台。
dsh plugin --profile web add @dickpy/dsh-imagegen
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-09-03. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.