dsh-video-director
面向 DeepSeek Harness 的项目级多模态视频导演插件。
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:chiphoton/DeepSeek-Harness-Video-Director#743913eeea6f0799631d3833c3d8693ff6ec8757
Video Director 以 dsh-video-director 发布，当前版本为 0.1.2。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。
简体中文 · Agent installation guide · Developer guide · Custom Nodes · Example Gallery
DeepSeek-Harness Video-Director
A visual, node-based video director for DeepSeek Harness.
Script · Prompt · Image · Audio · Video · Fully local workflows
DSH Plugin · ComfyUI · Ollama · MiniMax-H3 · Uncensored
DeepSeek-Harness Video-Director is a video-production plugin built for DeepSeek Harness. It turns script, image, audio, and video generation into a connectable canvas, with ready-to-use paths for ComfyUI, Ollama, OpenAI-compatible APIs, and Codex Plan.
Beginners can start with the built-in workflows instead of assembling every provider call by hand. Native Qwen3.8-27B and MiniMax-H3 paths make a fully local, uncensored, deployment-controlled multimedia pipeline possible when the selected models and runtime support it.
| Highlight | What you get |
|---|---|
| 🧩 Visual production graph | Connect text, images, audio, video, workflows, previews, and saved outputs on one infinite canvas. |
| 🚀 Beginner-ready workflows | Start from bundled image, H3 video, H3 audio, and prompt-enhancement workflows. |
| 🏠 Local-first generation | Use Ollama and ComfyUI on your own machine, including Qwen3.8-27B and MiniMax-H3 pipelines. |
| 🔌 Multiple providers | Mix Ollama, OpenAI-compatible endpoints, Codex Plan, and one logical ComfyUI backend in the same project. |
| 🎞️ Project-aware direction | Every Video Project keeps its own canvas, chat session, jobs, immutable assets, and provider choices. |
| 🛠️ Extensible by design | Import reviewed ComfyUI API workflows or package them as declarative Video Director Custom Nodes. |
Requirements: Git, Node.js ^22.19.0 or >=24, pnpm 11.7.0, and an installed dsh CLI. For a single-machine, fully local deployment, NVIDIA DGX Spark is the recommended environment, but it is not required; the exact hardware requirement depends on the workflow profile and model precision you select. Run the following from the directory where you want to keep the plugin:
git clone --branch uncensored --single-branch https://github.com/chiphoton/DeepSeek-Harness-Video-Director.git
cd DeepSeek-Harness-Video-Director
pnpm install --frozen-lockfile
pnpm run build
dsh plugin --profile web add "file:$PWD"
dsh web
dsh web opens the DSH Web UI. The plugin is installed persistently into the web profile, so later starts only need:
dsh web
If you run DSH from a sibling source checkout instead of an installed CLI:
# Run once in DeepSeek-Harness-Video-Director
pnpm install --frozen-lockfile
pnpm run build
# Then run in the sibling deepseek-harness checkout
pnpm dsh plugin --profile web add ../DeepSeek-Harness-Video-Director
pnpm dsh web
If Ollama or ComfyUI is not ready yet, give the agent installation guide to an agent. It inventories the exact models, ComfyUI workflows, and Custom Nodes used by this repository, installs only the selected workflow profile, and verifies the result. The default path installs both services locally; a separate path keeps cloud services bound to their remote loopback interfaces and reaches them through SSH local port forwarding.
Open Settings → Connections in Video-Director:
127.0.0.1:11434; install a model such as Qwen3.8-27B on the Ollama host, then select it in Video-Director.127.0.0.1:8188; install each workflow's required models and Custom Nodes on the ComfyUI host.You only need one working provider to begin. Video-Director does not silently install models, ComfyUI nodes, or external services.
Double-click empty canvas space, add nodes, and connect compatible handles:
Text → Prompt Enhancer → H3 Video → Preview → Save Output
Set the provider and workflow on each executable node, enter a prompt, then click Run. Use Save in the top bar to persist canvas edits; generation can run from the current unsaved canvas snapshot.
DeepSeek-Harness Video-Director node canvas
| Group | Nodes | Use |
|---|---|---|
| Inputs | Text, Image, Audio, Video, Sketch | Type, upload, paste, drop, or draw source material. |
| Workflows | Prompt Enhancer, Image Processing, H3 Video, H3 Audio | Generate or transform media with the selected provider and workflow. |
| Utilities | VRAM Trigger | Insert an execution barrier and optionally eject Ollama models or clear ComfyUI VRAM/cache. |
| Outputs | Preview, Save Output | Inspect results in the project or download them with an explicit filename. |
| Custom Node | Bundled and imported definitions | Run typed, reusable ComfyUI-backed nodes with compact primary and Advanced controls. |
Useful canvas gestures:
Bundled Custom Nodes include Qwen image editing, Z-Image Turbo, MiniMax-H3 text/image-to-video, reference-to-video, and Turbo/Standard H3 audio workflows. See custom_nodes/ for dependencies and safety notes.
The bundled comfyui-workflow-to-node Skill converts a trusted ComfyUI workflow into either:
Invoke it from a DSH conversation:
$comfyui-workflow-to-node
Convert /absolute/path/my-workflow-api.json into a reusable Video Director
Custom Node. Keep model and sampler controls in Advanced.
API-format workflows can be analyzed offline. Editor/UI workflows require exact metadata from the matching ComfyUI /object_info; the Skill stops instead of guessing when a mapping is ambiguous. Conversion does not submit the graph, install Python Custom Nodes, download models, or generate media.
The Skill is already prepared for plugin distribution: its project source lives under skills/comfyui-workflow-to-node/, package.json includes skills/ in the published package, and the Host strips the Skill's YAML front matter before registering its body with the DSH Skills service while retaining its local references and script.
The default Host data directory is ./.dsh-video-director, resolved from the directory where dsh web starts:
.dsh-video-director/
├── projects/<project-id>/project.json # canvas, settings, and recent jobs
├── assets/<asset-id>.<ext> # uploaded and generated media
├── assets/index.json # immutable asset metadata and hashes
└── workflows.json # imported workflow registry
Change dataDir in the plugin's DSH configuration if you need a stable absolute location. Save Output downloads a copy through the browser to the browser's configured download directory; the project-owned asset remains under dataDir/assets.
For configuration, architecture, transport behavior, recovery, security, limitations, and development commands, read the developer guide.
MIT. Model weights, ComfyUI, Custom Nodes, and external services retain their own licenses and terms.