SKILL.md
Video Generate
Generates a video from a text prompt using AI video generation models. Supports both text-to-video and image-to-video (animating a static image).
Command
agent-media video generate --prompt <text> [options]Inputs
| Option | Required | Description |
|---|---|---|
--prompt |
Yes | Text description of the video to generate |
--in |
No | Input image path or URL for image-to-video |
--duration |
No | Duration in seconds (6, 8, 10, 12, 14, 16, 18, 20; default: 6) |
--resolution |
No | Video resolution (720p, 1080p, 1440p, 2160p; default: 720p) |
--fps |
No | Frame rate (25 or 50; default: 25) |
--audio |
No | Generate audio track |
--out |
No | Output path, filename or directory (default: ./) |
--provider |
No | Provider to use (default: auto-detect) |
--model |
No | Model to use (overrides provider default) |
Output
Returns a JSON object with the generated video path:
{
"ok": true,
"media_type": "video",
"action": "video-generate",
"provider": "fal",
"output_path": "generated_abc123.mp4",
"mime": "video/mp4",
"bytes": 12345678
}Examples
Generate a video from text:
agent-media video generate --prompt "a cat walking in a garden"Generate a longer video with higher resolution:
agent-media video generate --prompt "ocean waves crashing on a beach" --duration 10 --resolution 1080pAnimate an existing image (image-to-video):
agent-media video generate --in portrait.png --prompt "the person waves hello and smiles"Generate with audio:
agent-media video generate --prompt "fireworks exploding in the night sky" --audioUse a specific provider:
agent-media video generate --prompt "a rocket launching" --provider replicateProviders
This action requires an external provider:
- fal - Requires
FALAPIKEY(uses LTX-2 model) - replicate - Requires
REPLICATEAPITOKEN(uses lightricks/ltx-video model)
The local provider does not support video generation.