DeepSeek Harness Plugin Hub

发布与管理完整 Harness Profiles,发现适合你的插件。

探索

插件目录环境预设文档中心动态

社区

发布插件联系我们报告问题

相关链接

Plugin Hub GitHubDeepSeek Harness 官方项目系统状态隐私说明
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

独立、非官方社区项目,与 DeepSeek 官方无隶属、授权或背书关系。

Videogen — DeepSeek Harness 插件(DSH Plugin)
← Plugins
V

dsh-videogen

Videogen

适用于 dsh Web GUI 的 AI 视频插件:多供应商视频生成渠道(OpenAI-compatible /v1/videos、Kling、MiniMax Hailuo、Volcengine ARK、自定义通用异步)、FFmpeg 视频处理(探测 / 帧提取 / GIF / 图像压缩)、轻量级分镜工作室(模板 + 镜头 + FFm

插件会安装到这里;不确定时保持 web。

npx -y @deepseek-ai/dsh plugin --profile web add github:shimingming520/dsh-videogen#f92bbb971e9d02966fb64b0320f00760a21acfb4
README兼容性版本

说明

适用于 dsh Web GUI 的 AI 视频插件:多供应商视频生成渠道(OpenAI-compatible /v1/videos、Kling、MiniMax Hailuo、Volcengine ARK、自定义通用异步)、FFmpeg 视频处理(探测 / 帧提取 / GIF / 图像压缩)、轻量级分镜工作室(模板 + 镜头 + FFmpeg 合成)、Agent 工具和侧边栏 AI 视频面板。

兼容性与来源证明

Videogen 以 dsh-videogen 发布,当前版本为 0.2.1。Plugin Hub 会校验它的 manifest,并保存精确安装来源,便于复现安装结果。

DSH 兼容范围
*
运行环境
web
发布来源
github
Registry 更新时间
2026/9/7

版本

0.2.1stable
2026/9/7

README

🎬 dsh-videogen

AI video generation + processing for DeepSeek Harness (DSH) — turn your DSH web GUI into a video studio: multi-vendor text-to-video / image-to-video channels, FFmpeg processing (frames / GIF / compress / concat) and a prompt-driven storyboard studio, from the sidebar panel or straight from the Agent.

简体中文 | English

✨ Features

  • Multi-vendor video channels in one place: OpenAI-compatible /v1/videos (Sora spec, gateway-friendly), Kling (text2video / image2video, JWT or Bearer), MiniMax Hailuo (task + file retrieve), Volcengine ARK (Doubao Seedance) and a fully custom generic async channel (submit URL, body template with {{prompt}} {{model}} {{image}} {{duration}} …, poll URL template, status/result paths) — add as many as you need
  • Keys stay local: API keys live in the local DSH settings document and every upstream call is proxied by the local host; the browser and the Agent never touch plaintext credentials
  • FFmpeg processing (host-side, reusing the dsh-video-tools feature set): info probe, frames extraction (evenly spaced or at a timestamp), video_to_gif (two-pass palette), image_compress, and concat with crossfade + background music
  • Storyboard studio: six bundled cinematic shot templates (brand sizzle, product launch, data story, vertical social, feature orbit, release spotlight) with {{variable}} slots; generate every shot through the configured channel, then compose the whole board into one video with FFmpeg
  • Agent tools: generate_video (text2video / image2video, wait or query by task_id), video_process, search_video_library, manage_storyboard — plus bundled session skills (/video:gen, /video:process, /video:storyboard)
  • History + resource library: every task recorded under ~/.dsh/dsh-videogen/, generated videos cached same-origin, optional auto-save to the library with full provenance
  • Prompt enhancement: rewrite a rough idea into a ready-to-generate description with the agent default model (no extra key)

📦 Installation

DSH host (Node ≥ 20) required.

dsh plugin --profile web add dsh-videogen

Local development install:

dsh plugin --profile web add /path/to/dsh-videogen

Restart dsh web — the sidebar shows the AI Video entry.

🚀 Quick start

  1. Open Settings → Plugins → AI Video
  2. Add a channel: pick a preset provider (+ Add provider) or + Add custom provider
  3. Fill in the API URL, API key (Kling accepts AccessKey:SecretKey for JWT signing), and the model catalog — use Fetch available models to import them
  4. Save, then open the AI Video sidebar panel: Generate tab (text2video / image2video, ✨ enhance), Process tab, Studio tab (templates → shots → compose) and Library tab

🎛 Channels

PresetSubmitPollAuth
OpenAI-compatiblePOST {base}/videosGET /videos/{id}Bearer
KlingPOST /v1/videos/text2video / image2videoGET …/{task_id}Bearer token or AccessKey:SecretKey → HMAC-SHA256 JWT
MiniMax HailuoPOST /v1/video_generation/v1 / image_to_video/v1task poll + /v1/files/retrieve/{file_id}Bearer (+ ?GroupId= on apiUrl)
Volcengine ARKPOST …/contents/generations/tasksGET …/tasks/{id}Bearer
Custom genericyour submitUrl + bodyTemplatepollUrlTemplateconfigurable header/scheme

🤖 Agent usage

ToolPurpose
generate_videoSubmit a text2video/image2video task (or query by task_id); waits by default and returns same-origin video URLs
video_processFFmpeg processing: info / frames / gif / compress / concat on URLs or ws:<workspaceId>/<path> files
manage_storyboardlist_templates / new / get / update_shot / attach_video / compose / generate_all
search_video_libraryFind and reuse previously generated/saved videos

Typical session commands (skills bundled with the plugin):

/video:gen        Cinematic shot of a neon-lit city, camera dolly-in, 9:16
/video:process    Extract 4 frames from the generated video
/video:storyboard Make a 15s product launch video for NEON X1

🔐 Security & data notes

  • API keys are stored in the local DSH settings document; requests are proxied by the local host (/api/dsh-videogen/*, loopback-only routes)
  • Generation consumes your upstream provider quota; video content is produced by the upstream model
  • History, cached videos, processing outputs and storyboard projects persist under ~/.dsh/dsh-videogen/
  • FFmpeg is resolved in order: FFMPEG_PATH → bundled @ffmpeg-installer/ffmpeg → system ffmpeg on PATH → ffmpeg-static; install via brew install ffmpeg if none exist
  • Prompt enhancement calls the LLM model you chose (default: agent default model) — no extra API key

🧪 Development

pnpm install
pnpm run typecheck   # tsc --noEmit
pnpm run test        # vitest run
pnpm run build       # tsdown → lib/index.js + lib/client.js

相关插件

继续浏览 vision-media 分类下经过校验的插件。

Tool Describe Image@linxin666/dsh-tool-describe-image面向模型的 describe_image 工具,用于 dsh Web GUI:通过在兼容 OpenAI 的端点调用视觉语言模型,为文本模型提供图像理解能力,以描述一张图像(本地路径、http(s) URL 或附件引用)。可热插拔 —Modlens@liustack/modlens面向仅支持文本的 LLM 的插件视觉能力,由免费的 Antigravity CLI 提供支持Deepseek Ivideodeepseek-ivideoiPolloWork HyperFrames Video Studio,以及 27 个可编辑视频模板,以原生 DeepSeek Harness 对话视图呈现。Image Gendsh-image-gen将类似 ChatGPT 的图像生成功能带到 DeepSeek Harness——支持 Gemini、OpenAI、Seedream、DashScope、本地 ComfyUI 等。
最新版
0.2.1
DSH
*
HMR
重启进程
Tree shaking
未声明可安全裁剪
解包体积
未提供
文件数
未提供
Surface
web
许可证
Apache-2.0
发布源
github
GitHub
★ 0
周下载
0
最近提交
2026/9/7
查看源码 ↗
README Badge

点击下方 Badge 复制 Markdown,粘贴到 README 即可。

这是你的 Plugin?认领权益 · 优先安全扫描

验证 package.json 声明的 GitHub 仓库,即可管理这个公开页面。认领后,Hub 会优先安排当前版本的安全扫描,并在通过后公开展示结果。

认领这个 Plugin →
报告问题
DeepSeek Harness Plugin Hub
ProfilesPlugins分类动态文档登录管理 Profiles
ProfilesPlugins分类动态文档登录