---

name: generative-media-skills description: | 41 workflow recipes for AI media generation covering Midjourney v7, Flux Kontext, Kling 3.0, Veo3, Seedance 2.0, and 100+ models. Covers image generation, video creation, audio (Suno), and social media content. Requires a muapi account and muapi-cli. triggers:

• "generative media"
• "Flux Kontext"
• "Kling 3.0"
• "Veo3"
• "SamurAIGPT" das: category: creative-3d upstream: "https://github.com/SamurAIGPT/Generative-Media-Skills" version: latest install: false

---

generative-media-skills

Catalogue stub — full platform: SamurAIGPT/Generative-Media-Skills

Account requirement

A muapi account is required. Sign up at muapi.ai before installing. The muapi-cli tool provides the runtime bridge to 100+ AI media models.

Install methods

Method	Command
Install all 41 skills	npx skills add SamurAIGPT/Generative-Media-Skills --all
Install individual skill	npx skills add SamurAIGPT/Generative-Media-Skills --skill <name>
Authenticate muapi	npx muapi-cli auth login

Decision tree

1. Do you have a muapi account?
   • No → direct user to muapi.ai to create an account first.
   • Yes → go to step 2.
2. Is muapi-cli authenticated? Run npx muapi-cli auth status.
   • No → run npx muapi-cli auth login
   • Yes → go to step 3.
3. Are the skills installed? Check ~/.design-agent-skills/skills/ for generative-media skill directories.
   • No → run the install command above.
   • Yes → invoke the relevant skill by name.

What it does

Provides 41 structured workflow recipes that orchestrate 100+ AI media generation models through the muapi platform. Covers text-to-image (Midjourney v7, Flux Kontext, DALL-E), text-to-video (Kling 3.0, Veo3, Seedance 2.0), music and audio generation (Suno), and social media content creation workflows. Each skill encapsulates prompt patterns, parameter tuning, and model selection for a specific media output type.

When NOT to use

• Without a muapi account — the skills require muapi-cli authentication
• fal.ai model workflows specifically — use fal-ai-skills instead
• Static image editing without AI generation