doany-ai

Choosing the right video model from dozens of RunComfy endpoints is hard. This skill routes your request to the best model for your intent and applies documented prompting patterns for quality results.

Choisir le bon modèle de musique IA pour chaque cas d'usage est complexe et coûteux. Cette skill achemine votre requête vers le modèle RunComfy optimal — ElevenLabs premium pour les campagnes phares, ACE Step économique pour la musique d'ambiance, ou des points de terminaison d'édition pour corriger des pistes existantes.

Picking the right AI image model for a task is hard, and the CLI syntax for each is different. This skill routes your request to the best model and ships the exact invoke command for text-to-image and image-to-image generation.

Users need to create lip-synced avatar videos but don't know which AI model to use. This skill routes requests to the right RunComfy model — OmniHuman, Wan 2-7, HappyHorse, or Seedance v2 — and generates the correct command in one step.

Creating music with AI is expensive and limited to text-to-audio. This skill provides four ACE Step endpoints for generation, inpainting, and outpainting at approximately 27 times lower cost than alternatives, with full control over tag-driven composition and multilingual vocals.