Use cawcut-generate to create an image or video from a prompt and optional media. It first checks whether an available CawCut App matches your request. If a suitable App is found, the skill can route you to that workflow; otherwise it continues with direct generation.
1. What cawcut-generate Does
Describe the image or video you want and provide any references the selected mode needs. The skill supports direct generation, image editing, image-to-video, and video generation using image, video, or audio inputs.
1.1 Choose a Generation Mode
The available modes include:
| Mode | Use it to |
|---|---|
| Text to image | Create an image from a written prompt. |
| Text to video | Create a video from a written prompt. |
| Image to image | Edit or restyle one or more reference images. |
| Image to video | Animate a still image. |
| Mixed media to video | Create a video using image, video, or audio inputs, alone or together. |
Which models and settings are available can change. The skill checks the current options for the selected mode before generation.
A request can include:
- The media type: image or video.
- The subject, action, or story.
- The visual style and mood.
- The camera angle, movement, or composition.
- The aspect ratio, duration, or intended platform.
- Reference images, storyboards, or existing assets.
- Details that must remain consistent or should be avoided.
2. Run the Command
Enter the command in a compatible AI-agent chat, followed by your request. The skill checks available Apps first and may offer a matching App before direct generation.
/cawcut-generate Create a photorealistic 16:9 image of an iced coffee on a weathered wooden table beside the ocean at sunrise.You can also enter only the command and let the skill ask what you want to generate:
/cawcut-generate3. Prompt Examples
Use these examples as starting points and replace the details with your own.
3.1 Generate an Image From Text
/cawcut-generate Create a photorealistic 16:9 image of an iced coffee on a weathered wooden table beside the ocean at sunrise. Use soft natural light, shallow depth of field, and a premium lifestyle-advertising look.3.2 Generate an Image From a Reference
Attach the reference image, then enter:
/cawcut-generate Create a 4:5 product image using the attached cup as the exact product reference. Place it on a bright coastal cafe table, keep the logo and cup shape consistent, and use warm morning sunlight.3.3 Generate a Video From a Concept
/cawcut-generate Create a 10-second 9:16 video of a hand picking up an iced coffee beside the ocean. Begin with a close-up of condensation on the cup, then pull back to reveal the beach. Use smooth camera movement and an energetic commercial style.3.4 Generate a Video From a Storyboard
Attach the storyboard and references, then enter:
/cawcut-generate Create a video from the attached storyboard. Use the attached character and location images as visual references, keep the character's appearance consistent across every shot, and follow the camera directions shown in the storyboard.3.5 Generate a Video From Media References
For a video request that uses more than one kind of source, attach the relevant files and state what role each one should play:
/cawcut-generate Create a short video using the attached product image as a visual reference, the attached clip for movement, and the attached audio as the soundtrack.4. Write More Effective Prompts
For more predictable results:
- State whether you want an image or video.
- Describe the main subject and action first.
- Include the aspect ratio and video duration when relevant.
- Explain how each attached reference should be used.
- Identify details that must remain consistent.
- Mention anything CawCut should avoid.
5. Review Your Results
When generation finishes, the skill returns result links and saves the media to your local Downloads folder. It also reports the credits used and your remaining balance. It then hands the local result files to cawcut-vn, which can offer to import them into VN or use them in a planned editing workflow.