On Meta in 2026 the algorithm finds the audience; **the creative is the targeting**. The lever a fashion team actually controls is what the first three seconds say and show. This skill turns "make more ads" into a test that produces an answer.

The whole morning in one ask: the user names a topic, and a published 9:16 short comes out the other end, about something that is actually news today and has not been covered before.

A rejected ad costs a day; repeated rejections put the ad account at risk, which costs far more. Fashion trips the same few rules over and over, so they can be caught before submission.

Peak days are when CPMs are highest and everyone is shouting. The shops that win them gathered their audience (video viewers, engagers, add-to-carts) and tested creatives **in the weeks before**.

Turn a topic into *evidence on screen*: the real announcement page, and footage that shows what the narration is describing. Without this a video is text slides read aloud.

A list of facts gets watched until the viewer has the gist. A story gets watched until it ends, because the viewer wants to know how it ends. This skill is `create-video` with one change: **the script is written on an arc first, and the frames are picked by the beat they carry, not by how they look.

The user says a topic. You come back with the two or three things worth making a video about *today*, each with a reason it is worth making today. Not a search-results dump — a shortlist you would defend.

UGC-style video from real people is the format fashion ads keep returning to: a real body, a real room, someone turning around in the mirror. It only works as an **ad** if the brief gets the first 3 seconds and the rights right.