Video to Previs

Shot2Previs converts a video shot into a temporally consistent previs (previsualization) reference — a clay-shaded 3D depth render with multi-person pose skeletons that captures the shot's camera movement, framing, and subject blocking without reproducing any pixels. Pass a video URL, poll the request, download the previs mp4; feed it to a generative video model (Seedance, Wan-VACE, LTX) as a structural reference. Optional green-screen mode isolates the performer.

You can ask Video to Previs to handle video generation. Powered by Pixazo via the Pixazo API gateway.

How to use this skill (agent instructions)

When this skill triggers, follow this script. The user should only ever have to give you their API key once.

Step 1 — Make sure you have a Pixazo API key

Check for the API key in this order:

1. PIXAZO_API_KEY environment variable
2. ~/.pixazo/api-key file (one line, just the key)
3. The current conversation (user pasted it earlier)