Assemble an iMessage notification-cascade video ad (≈14s, 9:16) from a plate of a phone lying face-up (setting chosen by the user) + 3–5 messages — authentic Apple Messages banners composited in PIL (SF Pro text, green Messages icon, warm translucent-greige fill, soft shadow) spring in one-by-one at the BOTTOM and push the stack UP, a right-aligned Show-less/X pill rides above, the X clears the stack, then a serif end card resolves. FREE assembly (PIL + ffmpeg); the recipe supplies the per-brand plate, notifications, and end-card config and gates the paid plate-clean/music calls to their own capabilities. Use for the imessage-notification-cascade format.
npx gooseworks install --all # then, in Claude Code, Cursor, or Codex: /gooseworks use the render-imessage-cascade skill
The free, deterministic renderer for the imessage-notification-cascade video ad format — the viral iOS trend where a phone lies face-up (the source: on a desk) and Apple Messages notifications STACK IN one after another. The signature mechanic is the bottom-up push: each new banner springs in at the bottom (nearest the phone) and shoves every existing one UP a row; the iOS grouped "⌄ Show less / ✕" pill rides above the stack; the ✕ clears the stack; then a serif end card resolves.
This is a DETERMINISTIC composite — no generative video of the UI. Authentic
iMessage banners are drawn in PIL and animated in FFmpeg over a Ken-Burns plate, so
the notification text + wordmark stay pixel-crisp (a video model would smear type).
The template recipe supplies the per-brand plate, notifications, and end_card
config and gates the only paid steps — cleaning the plate (→ create-image-fal) and
the music bed/pop (→ create-music-elevenlabs) — to their own capabilities. This
capability itself makes no paid calls.
The creative calls are the user's, asked by the format recipe before any paid step; this renderer draws whatever the config holds.
notifications[]. The demo used inbound leads.config.audio.
With no bed, compose.py plays the pops + swoosh over silence; --no-audio is fully silent.
The demo used a soft music bed.scripts/build_assets.py — draws the assets from config: nb-1..N.png (authentic
banners — green Messages icon, warm translucent-greige fill, soft box-shadow,
title/body/NOW/handle), pill.png (right-aligned "⌄ Show less / ✕"), endcard.png
(serif CTA + wordmark lockup + url). Fonts are load-bearing: SF Pro (SFNS.ttf)
via set_variation_by_name for the banner title/body/NOW/handle (Arial fallback),
Times/serif for the end-card CTA. Do NOT swap in Helvetica/Arial as the primary —
the banners must read as the real iOS system font.scripts/compose.py — Ken-Burns push-in on the plate → each banner springs in at the
BOTTOM while later arrivals push the stack UP (FFmpeg overlay y expressions) → pill
rides above → ✕-clear swipes the stack up + fades → serif end card fades in → optional
audio (optional bed + pop per arrival + a free FFmpeg swoosh on the clear) → encode h264 + aac.scripts/config.example.json — the shape of the brand config the recipe binds.W=1080 H=1920, SIDE=135 → banner width BODY_W=810, BANNER_H=176, PAD=60,
row pitch H=214, bottom anchor YB=1200. Icon ~100px at a ~24px left inset; body
text starts ~150px from the banner's left edge. Change one, change both.
watch (QC the final master). The recipe gates create-image-fal (plate clean) and
create-music-elevenlabs (bed/pop) — both paid, proxy-routed, billed to the Ads agent.
config.audio.scripts/build_assets.py — draws the assets from config: nb-1..N.png (authenticscripts/compose.py — Ken-Burns push-in on the plate → each banner springs in at theFull video production sequence with script review, actual ingredient choices, controlled generation, editing, evidence-based quality review, polish, captions and delivery. A host binding supplies project storage, authentic approvals, provider access and billing.
Build a vox-pop street interview video ad. An interviewer with a handheld mic asks passers-by one question about the brand's product, they give blunt wrong guesses, one gives the real answer, and the cut lands on a branded end card. Generates the takes through the GooseWorks fal proxy (Seedance 2.0 with native voice), then grades, re-cuts, captions and gates them locally. Use for the street-interview format.
Write the words of a short-form video ad (voiceover, dialogue, chat bubbles, on-screen lines) the way performance creative teams do instead of from a blank page. Builds the script from the buyers' own words, the beat sheet of an ad that already works and three deliberately different angles, filters them with a rule check and a second non-Claude model, and takes the strongest into the review with the other two as one-line swaps. Use it in every video ad run before any paid step, and whenever the user asks to write, rewrite or improve a video ad script or says a script sounds generic or AI-written.
An iMessage video ad shows a text conversation on an iPhone screen. Here is how a single skill teaches Claude to build one, with sound and an end card.