dsh-image-gen poster

# 🎨 dsh-image-gen **Complete conversational image capabilities for DeepSeek Harness: text-to-image, image-to-image, multi-image references, continuous editing, local ComfyUI, and gallery management.** [![npm version](https://img.shields.io/npm/v/dsh-image-gen.svg?style=flat-square&color=blue)](https://www.npmjs.com/package/dsh-image-gen) [![DSH Plugin](https://img.shields.io/badge/Plugin%20For-DeepSeek%20Harness-6366f1?style=flat-square)](https://github.com/deepseek-ai) [![npm downloads](https://img.shields.io/npm/dm/dsh-image-gen.svg?style=flat-square&color=emerald)](https://www.npmjs.com/package/dsh-image-gen) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg?style=flat-square)](LICENSE) **English** | [简体中文](README.md)

💬 Prompt your DeepSeek Harness Agent to install:

```text Install the image generation plugin by running: pnpm dsh plugin --profile web add dsh-image-gen@latest ```

_{(Or run manually in terminal: pnpm dsh plugin --profile web add dsh-image-gen@latest)}

After installation, enter your API key or configure local ComfyUI in DSH Settings, then tell the Agent:

```text Draw a cyberpunk cat on a neon street in a rainy night. ```

The Agent will automatically generate images and can also continue editing the previous image directly.

Initial image generation conversation

Continued image editing conversation

--- ## 💡 What Problem Does It Solve? **`dsh-image-gen` is an open-source image generation and editing plugin built specifically for DeepSeek Harness (DSH).** DeepSeek Harness empowers agents to use tools for various tasks. This project adds native **multimodal image generation capabilities**: ```mermaid graph LR A[User Prompt] --> B[DeepSeek Harness Agent] B --> C[generate_image / edit_image] C --> D[Gemini / OpenAI / Seedream / DashScope / ComfyUI] D --> E[Image Data] E --> F[In-chat Conversation Stream] ``` --- ## 🚀 Quick Install & Usage ### 1. Install Plugin In your DeepSeek Harness workspace root: ```bash # Recommended: Install or upgrade to latest release pnpm dsh plugin --profile web add dsh-image-gen@latest # If dsh is installed globally: dsh plugin --profile web add dsh-image-gen@latest ```

🛠️ Alternative Installations (Git Repo / Local Dev)

```bash # Method B: Install directly from GitHub pnpm dsh plugin --profile web add git+https://github.com/shanliuling/dsh-image-gen.git # Method C: Clone & install locally git clone https://github.com/shanliuling/dsh-image-gen.git pnpm dsh plugin --profile web add ./dsh-image-gen ```

### 2. Configure Provider & Workspace Settings Open DSH Web (`http://localhost:3080`): 1. Navigate to **Settings → Plugins → Image generation**. 2. Select a Provider. Enter an API key for cloud providers; for local ComfyUI, enter its URL and import an API Format Workflow JSON using `{{prompt}}` for the prompt and optional `{{seed}}` for the seed. 3. Optionally enable **Save to workspace** (enabled by default) and specify subfolder, then click **Save**.

Settings preview

### 3. Generate Images in Chat Simply prompt in the chat input: ```text Generate a minimalist modern architecture living room illustration. ``` The Agent will call `generate_image` and return the image directly in the conversation stream. ### 4. Browse Native Image Gallery Click the **`[Gallery]`** Tab in the top navigation bar to browse and search all generated images across conversations:

Gallery preview

--- ## ✨ Key Features - 💬 **In-Chat Image Generation**: Tell your Agent what you want to draw without switching tabs or tools. - 🖼️ **Native Image Gallery**: Dedicated "Gallery" tab automatically collecting all generated images with keyword search, provider filtering, single-item deletion (with confirmation and persistent tombstones), and quick copy/download. - 🔍 **Interactive Image Tools**: High-res fullscreen preview, one-click copy to clipboard, local download, and open in new tab. - 🎨 **Multi-Provider Support**: Supports Google Gemini, OpenAI Images, OpenAI Compatible API, ByteDance Seedream / Volcengine Ark, Aliyun DashScope (Wanx / Qwen-Image), and local ComfyUI workflows. - 🔑 **BYOK (Bring Your Own Key)**: Uses your own API keys managed securely by DSH credentials service with write-only protection. - 🖼️ **Durable Session Persistence**: Images integrate with DSH Attachment and conversation lifecycle, preserved across reloads. - 💾 **Workspace File Output**: By default each generated image is also written as a file into the current session workspace; the tool result carries the absolute file path. Can be disabled or re-pointed in settings. - ⚙️ **Native Web Settings**: Configure providers, keys, models, and endpoints directly in DSH settings. --- ## 📦 Supported Providers | Provider | Default Model | Default Endpoint / Base URL | | :--- | :--- | :--- | | **Google Gemini** | `gemini-3.1-flash-image` | `https://generativelanguage.googleapis.com/v1beta/interactions` | | **OpenAI Images** | `gpt-image-2` | `https://api.openai.com/v1` | | **OpenAI Compatible** | Custom | Custom Base URL | | **ByteDance Seedream / Volcengine Ark** | `doubao-seedream-5-0-260128` | `https://ark.cn-beijing.volces.com/api/v3` | | **Aliyun DashScope / Wanx** | `wanx2.1-t2i-turbo` | `https://dashscope.aliyuncs.com/api/v1` | | **Local ComfyUI (text-to-image only)** | Imported API Format Workflow | `http://127.0.0.1:8188` | --- ## 🛠️ Local Development ```bash git clone https://github.com/shanliuling/dsh-image-gen.git cd dsh-image-gen pnpm install pnpm run typecheck pnpm run test pnpm run build pnpm run pack:check ``` --- ## 📄 License Open-sourced under the [MIT License](LICENSE). If this plugin helps you, feel free to give it a ⭐️ **Star**!