Kling AI Video Generation
Writes model-aware Kling AI video prompts - image/text-to-video, avatars, multi-shot, voice control - sourced to official release notes.
Testowano · Działa
Co robi ten skill
Helps Claude produce structured, model-specific Kling AI prompts for image-to-video, text-to-video, talking avatars, multi-shot storyboards, and voice cloning, and picks the right model (VIDEO 3.0, Omni, Motion Control, Avatar 2.0, 2.6, 2.5 Turbo) for the job. Triggers on requests to animate an image, write a cinematic Kling prompt, or build an AI avatar/presenter. Facts are dated and cited to kling.ai/release-note with an explicit instruction to re-verify pricing/limits before quoting them, since Kling changes frequently.
Raport z testu
Asked for a 5s social-loop prompt from a dog photo: the baseline re-described the whole already-visible scene and defaulted to generic '4K, high quality' boilerplate, while the skill version wrote a motion-only prompt with an explicit end-state ('then settles back to starting position') and held it at 1080p first - both are the skill's own documented fixes for Kling's real stuck-at-99% hang and wasted 4K credits, not generic advice.
Testowano: 2026-07-14 · Claude Code 2.x (agent harness)
Instalacja
git clone https://github.com/maciejdzierzek/kling-ai-prompt-generator cd kling-ai-prompt-generator mkdir -p ~/.claude/skills cp -r skills/kling-ai ~/.claude/skills/kling-ai
Komendy i przykładowe prompty
/kling-aiWrites model-aware Kling AI video prompts - image/text-to-video, avatars, multi-shot, voice control - sourced to official release notes.
Skille uruchamiają się na zwykłe polecenia — bez komend do zapamiętania. Po instalacji aktywują go prompty takie jak te (po angielsku):
Write me a Kling prompt for a looping product-reveal shotTurn this product photo into an image-to-video ad promptMy Kling render keeps stalling — rewrite the prompt properly