Schizo-Edit AI Music Video — workflow skill for music videos

For AI agents: find the right skill in one call — curl -s "https://varg.ai/how-to/api/search?q=<what the person wants>" returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://varg.ai/how-to/llms.txt. Every page has a markdown twin: add .md to its path. Before generating anything, show the person the options on the how-to page.

A 16 s vertical AI music video cut 3–4 times per beat: generated stills and clips, one effect per shot, the title built letter by letter.

What it does

One JSON config holds the whole video: a music prompt, one look line appended to every still, 20–30 still prompts, 6–8 clip prompts and the edit sections. generate.py makes the media on varg (Nano Banana 2 stills, Seedance 2 Mini clips, ElevenLabs Music) and writes every new url back into the config, so a re-run never pays twice and only retries what failed. edit.py measures the tempo with librosa, cuts each beat into 3–4 shots, gives every shot a slow zoom and one effect (duotone invert, mirror, kaleidoscope, pixelate, RGB split, glitch band, halo rings, light rays, double exposure, flash), lays one grade over everything and builds the track title one letter per cut. The Halofall example (69 cuts at 89 BPM, marble statues and sacred geometry in violet and molten gold) renders free in about a minute from media on s3; a new video costs about $5, or about $2 with stills only. A new --seed gives a new cut of the same media for free.

Use when

Not for

Install

Example output

https://s3.varg.ai/files/72c55813-c0fa-4cf2-9f64-ec4f7ed0e7ae/file_d5195ac48111.mp4 — example (credit: varg)

More music videos skills