ugc-asmr-sensory-format
Installation
SKILL.md
ASMR and Sensory Format
The best fit for generated video in the entire UGC set: no face, no dialogue, no experience claimed, no identity to keep consistent. Just texture, motion and sound.
It also has the highest completion rates in short-form, and it's the format where "commercial polish" is least penalised — sensory content is allowed to look good, it just has to look tactile.
Why this format suits generation
| Constraint elsewhere | Here |
|---|---|
| Face consistency across clips | No face |
| Lip sync at 3 credits per second | No dialogue |
| Fabricated personal experience | No claims made |
| Creator persona design | Not needed |
| Disclosure of a synthetic person | No person depicted |
What's left is the thing models are genuinely good at: material behaviour under slow, simple motion. Which means the cost per usable clip is the lowest of any UGC format, and the iteration loop is fast.
The AI-content label still applies to generated footage. The rest of the compliance surface mostly doesn't. See ugc-disclosure-compliance.