Wraps the HappyHorse 1.0 text-to-video model through RunComfy's CLI. Currently ranks #1 on the Artificial Analysis Video Arena and ships with in-pass synchronized audio, which is rare: most models require you to add sound in post. Native 1080p, up to 15 seconds, supports six prompt languages. The skill documents when to route to sibling models instead (Wan 2.7 for audio-driven lip sync, LTX 2 for speed). Best for multi-shot stories where you need the same character across clips or when you want ambient sound and dialogue baked into the generation. Requires a RunComfy account and the CLI installed.
npx skills add https://github.com/agentspace-so/runcomfy-agent-skills --skill happyhorse-1-0