#113 by installsResearch & ML
ai-video-generation
Generate AI videos from text or images using RunComfy's full model catalog with smart routing
383K
3.5K
53
May 13, 2026
Install
npx skills add https://github.com/prime-skills/runcomfy-agent-skills --skill ai-video-generationWhat it does
Generate cinematic AI videos via the RunComfy CLI, which intelligently routes your request across HappyHorse, Kling, Veo, Seedance, Wan, Hailuo and Dreamina. Use it whenever you need text-to-video, image-to-video, video extension, or character-consistent multi-shot content. The skill handles model selection for your intent—Arena-quality output, audio sync, motion speed, or duration—and ships the exact prompt patterns and CLI invokes for each.
- Route text/image prompts to the right model: HappyHorse for Arena #1, Kling for cinematic, Veo for extend, Wan for lip-sync
- Text-to-video (t2v), image-to-video (i2v), and video extend with documented prompt patterns per model
- One `runcomfy run` command per generation; handles auth, queuing, and delivery
- Picks fastest path for sub-15s clips, longest duration, or multi-shot character identity based on user intent
