SamurAIGPT/Generative-Media-Skills ? reverse-engineered prompt

Build me a set of AI agent skills for generating and editing media, mostly for Claude Code, Cursor, Gemini CLI, and OpenCode.

I want it to let an agent create high quality images, videos, and audio from simple prompts, and also work with local files like photos, videos, faces, and audio that get uploaded automatically for processing. It should include a small core layer for the basic actions, plus a bigger library of ready to use creative workflows like product ads, logo animation, cinematic scenes, shorts, and other common media tasks.

Make it easy for an agent to run commands, get structured JSON back, filter results cleanly, and open the finished media in the viewer when asked. If it makes sense, add an MCP server so the tools can be used directly inside supported agents. Keep the whole thing focused on good creative defaults, professional looking outputs, and simple setup, and look up current docs online if you need to.