@icedcola/dsh-provider-veark火山方舟 Coding Plan LLM provider for DeepSeek Harness — provider-scoped PDF、图片与 Responses API 流式支持
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:IcedWatermelonJuice/dsh-provider-vearkDiscover The Best DSH plugins
1447results
Find image understanding, screenshot analysis and visual interaction plugins; compare inputs and usage.
@icedcola/dsh-provider-veark火山方舟 Coding Plan LLM provider for DeepSeek Harness — provider-scoped PDF、图片与 Responses API 流式支持
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:IcedWatermelonJuice/dsh-provider-veark@lhmd/dsh-director-toolkitDeepSeek Harness plugin that turns a 3D idea into a shootable scene brief and publish-ready showcase copy.
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:lhmd/dsh-director-toolkit@local/dsh-project-file-explorerDeepSeek Harness 项目文件浏览器:右侧停靠文件树 + 点击文件在主区打开预览标签(代码/文本/图片/音视频/PDF),新增工作区自动停靠,窄屏响应式抽屉。
@lyhue1991/dsh-soupDSH 项目资源管理器:右侧 details 列文件树(并列网格、宽度对齐宿主 details 合同 300-520px),有 Session log 时在其右侧、空白新会话在输入框上方右端提供 📁 切换按钮;支持多选、双击展开/打开、右键菜单(打开/系统打开/重命名/废纸篓/新建/复制路径)、拖拽移动、上传。文件「打开」在会话区新增「文件」标签页(原生 conversation.view,与对话/轨迹同级):文本可编辑 ⌘S 保存、图片预览、二进制提示。附带模型吞吐徽标:输入框正上方显示「正在等待模型...」与彩色 t/s 徽标(数据来自 llm/stream 真实流)。附带 GoalBar 多行展示:把原生单行截断 GoalBar 放开为多行完整显示,编辑用 textarea,工具与数据仍走原生 goal。
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:lyhue1991/dsh-soup@mackwan84/dsh-image图像生成 capability 的 Service Definition:UI 草图生成/编辑的抽象契约与错误码
@mackwan84/dsh-image-dashscope阿里云百炼图像生成 Provider:qwen-image 3.0 与 Wan 2.7 系列的文生图、参考图 I2I 和异步任务调用
@mackwan84/dsh-image-volcengine火山方舟图像生成 Provider:doubao-seedream 4.0-5.0 系列(文生图 + 参考图 I2I + 指令编辑)的同步调用
@mackwan84/dsh-tool-ui-mockupui_mockup 工具 Consumer:研讨阶段生成 UI 线框图/高保真草图,落盘设计稿并提示确认后锁定设计规格;含客户端卡片、图片路由与 i18n
@max-null/dsh-captureSSiD (思灵) quick screenshot capture: tray/hotkey → fullscreen box-select overlay → cropped image into the current conversation composer
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:Max-Null/dsh-capture@maxwell-feng/dsh-tesseract-ocrdsh plugin: recognize attached images locally with Tesseract OCR and send only the recognized text to the model. Text models never receive image bytes; vision passthrough is opt-in.
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:maxwell-feng/dsh-tesseract-ocr@pirate-608/dsh-after-effectsAdobe After Effects automation through ae-mcp
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/adobe-after-effects@pirate-608/dsh-autocad-mcpAutoCAD and headless DXF automation through a text-first MCP preset
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/autocad-mcp@pirate-608/dsh-comfy-local-toolsLocal-first ComfyUI generation and workflow automation through first-party MCP and CLI tools
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/comfy-local-tools@pirate-608/dsh-modlensText-first DSH vision bridge using Codex CLI, OpenAI-compatible endpoints, and Ollama
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/modlens@pirate-608/dsh-photoshopAdobe Photoshop automation through a local COM MCP bridge
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/adobe-photoshop@pirate-608/dsh-plugin-kitManaged preset and MCP approval primitives for pirate-608 DSH plugins
@pirate-608/dsh-premiereAdobe Premiere Pro automation through a bundled CEP MCP bridge
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/adobe-premiere@pirate-608/dsh-renpy-visual-novel-devRenPy development, preview, testing, and asset-integration workflows
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/renpy-visual-novel-dev@pirate-608/dsh-solidworks-automationSolidWorks COM and MCP automation workflows
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/solidworks-automation@pirate-608/dsh-unity-mcpA text-first Unity MCP agent preset and skill pack for DeepSeek Harness
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:pirate-608/dsh-multi-tools#6ac433e02e836f77218737a67747a8c32f376021&path:plugins/unity-mcp@yogemow/deepseek-prism-dshDeepSeek Prism for DeepSeek Harness (DSH): 纯文本模型图片附件降级 + 技能运行时注册 + 设置界面(视觉 API 密钥掩码输入与模型选择)。Image-attachment degradation, runtime skill registration, and settings UI (masked vision API key + model selection).
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:YOGEMOW/DeepSeek_Prism#f1147825809a75cdc28553e4c61df24a15f7b4f1&path:packages/plugin-dshagnes-mediaRegisters generate_image and generate_video tools for Agnes AI media models (agnes-image-2.5-flash, agnes-video-2.5-flash) in DeepSeek Harness. Supports multi-image reference via reference_image_urls.
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:LittleBeaverStudio/agnes-mediaawesome-dsh-background-pluginDSH Web background settings plugin with local image upload and live preview.
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:leavestring/awesome-dsh-background-plugindeepseek-harness-desktopDeepSeek Harness 桌面版 —— 内置官方 dsh 运行时与本地文件/视觉插件的自包含 Electron 桌面壳(macOS / Windows)
Use a feature keyword or category, then sort by recent updates or GitHub stars.
Read the description, AI analysis and original README to understand what a plugin does.
Review compatibility, dependency ranges and instructions, then copy the available install command.
DSH Hub is a community plugin directory for DeepSeek Harness, bringing together descriptions, source links and installation instructions to help you discover and compare extensions.
No. Detail pages show recorded compatibility and dependency ranges. Unknown does not mean compatible; check the project documentation before installing.
Open its detail page and check the installation section. Copy the command into your Harness environment when available, or follow the project README.