Go grab yourself an google gemini key and we're going to teach this agent to make files and videos.
Add 2 skills, one to generate images and the other to generate videos. Use the "npx @the-focus-ai/nano-banana" command, and look through how it works before writing out the skill. do some research first on the best ways to prompt these models, and incorperate your findings in the skill
Lets test:
make an image of diagraming the STEPS.md file
Here we go:

make a 720p 8 second video animating that image
Skills are super useful as you can see -- we are teaching it how to use a CLI in this case, and it comes with very very additional tokens in the context window.