AI Agent Hub
Back to plugins
⚙️

dsh-media-gen

Workflow Updated 2026.08.17

Run the following command in DeepSeek Harness:

dsh plugin install AtlasCloudAI/dsh-media-gen

Paste the following prompt into your AI chat to install this plugin:

Install by running dsh plugin install AtlasCloudAI/dsh-media-gen in your DeepSeek Harness profile; the source repository is https://github.com/AtlasCloudAI/dsh-media-gen

About this plugin

Generating image, video, audio, or 3D content is rarely bottlenecked by the model itself. The real friction is the repetitive chain of model comparison, shot planning, cross-model prompt adaptation, and parameter assembly that you redo by hand on every project. dsh-media-gen compresses that chain into three composable Skills: the atlas-cloud Skill recommends the right Atlas Cloud model and outputs validated parameters with a REST, CLI, or MCP execution path; the seedance-2-5-skill focuses on multi-shot Seedance video continuity, producing a storyboard and ready-to-use prompt; the universal-video-prompt-skill compiles one model-neutral video brief into model-specific prompt variants. Together they eliminate the need to hand-wire every API request.

If you also want DSH to trigger generation, upload assets, or poll jobs directly, the bundle includes an optional atlascloud-mcp@1.5.0 bridge. It is disabled by default; once you opt in via a profile override, DSH gains callable tools for model lookup, media upload, generation submission, task polling, and account usage checks—no additional scripting required.

This plugin suits creators working across multiple generation models, teams that reuse the same media brief across projects, and engineers who want a standardized, auditable media-production pipeline inside DeepSeek Harness. No Atlas Cloud credential is needed to use the planning and comparison capabilities; a key is only required when you actually submit a billable generation call.

Use Cases

  • Comparing multiple generation models and producing a storyboard with parameters for a multi-shot product video
  • Compiling one model-neutral video brief into model-specific prompt variants across different generators
  • Standardizing media production pipelines inside a DeepSeek Harness workflow with an auditable trail

Best For

  • Creators juggling multiple image, video, or 3D generation models
  • Cross-functional teams that reuse the same media brief across projects
  • Engineers who want a standardized, CI-verifiable media production pipeline inside DeepSeek Harness