poyo-kling-3-0

What This Skill Does

The poyo-kling-3-0 skill provides a powerful interface for interacting with the Kling 3.0 video generation engine via the PoYo API. It enables users to generate high-fidelity, creative videos by handling complex configurations such as single-shot and multi-shot workflows. The skill acts as an abstraction layer for the https://api.poyo.ai/api/generate/submit endpoint, allowing you to seamlessly manage model selection (between standard and pro tiers), aspect ratio definitions, and advanced features like sound generation, element referencing, and frame-based conditioning. Whether you are building a story through multi-prompt sequences or generating a single cinematic shot with reference frames, this skill handles the heavy lifting of payload construction and request submission.

Installation

To integrate this skill into your environment, run the following command in your terminal: clawhub install openclaw/skills/skills/coolhackboy/kling-3 Ensure you have the necessary credentials configured for the PoYo API before attempting to run submission scripts.

Use Cases

• Cinematic Storytelling: Use the multi-shot functionality to link together a cohesive narrative sequence with multiple prompts.
• Commercial Advertising: Utilize the kling-3.0/pro model to generate high-quality product showcases with controlled aspect ratios like 16:9 or 9:16 for social media.
• Animated Content: Leverage reference images for start/end frames to maintain character or environmental consistency across a video generation.
• Creative Sound Design: Enable the sound parameter to generate videos that automatically include audio-visual alignment, perfect for short-form content.

Example Prompts

1. "Generate a 10-second cinematic video of a futuristic cyberpunk city using kling-3.0/pro in 16:9 aspect ratio."
2. "Create a multi-shot video sequence: shot 1 is a close-up of a coffee machine, shot 2 is a steam rise, shot 3 is a barista pouring latte art. Use standard Kling 3.0."
3. "Make a video using this reference image [URL] as the start frame. The scene should show the landscape slowly transforming into a dreamscape during sunset, include sound."

Tips & Limitations

• Model Selection: Always prefer kling-3.0/pro for production-grade, high-quality results. Reserve kling-3.0/standard for rapid prototyping to save costs.
• Reference Frames: When using kling_elements or start/end frames, ensure your image URLs are publicly accessible, or the API request will fail.
• Multi-Shot Structure: Keep your multi_prompt segments distinct. If the AI struggles with transitions, ensure your prompts explicitly describe the temporal connection between shots.
• Polling: Always note the task_id returned by the script. Use the status-check endpoint periodically, as video generation for Kling 3.0 can take several minutes depending on the queue length.