Skip to main content
contenthero-generate turns an idea into media. It is execution only: you bring the prompt, the skill produces the asset and knows how to prompt each model on our roster.

​
What it does

  • Image, video, audio, reference board, and lip-sync generation.
  • Output-id chaining: feed a generation’s output id straight into the next as a reference (a start frame, an image input). No re-upload.
  • Cost preflight: preview any cost for free before committing.
  • Model prompt craft: the skill carries durable, model-specific prompting guidance, so a prompt is tuned to the model that will run it.

​
When the skill engages

When you ask your agent to make an image, video, audio, board, or lip-sync clip, or to chain one into another. It picks the transport (MCP or CLI), previews cost on large or batched generations, and keeps the output ids so it can chain.

​
How it grounds

For on-brand visuals, the skill can pull visual details the brand has documented before generating, via the brand skill. Identity (consistent characters across generations) flows through avatars and reference boards as typed references.

​
The full guide

This page is the summary. The skill’s own SKILL.md and its references/ (model catalog, prompt craft, identity, media inputs, chaining, troubleshooting) are included with the install. The corresponding product guide is Generate.