/render-cgi-sizzle

name: render-cgi-sizzle
description: Assemble a 3D-CGI app sizzle — nano-banana CGI plates (blank-glow floating phone + a CGI studio with rim-light + placeholder burst shapes; studio look and phone finish come from the recipe choices) plus PIL compositing of the REAL App Store screenshots onto the bezel + burst-out overlays, driven by Kling 3.0 i2v steady-float per beat with a per-beat Ken-Burns FFmpeg push-in fallback when Kling garbles the UI, then VO/music mix (sidechain duck, loudnorm) + 1.15x speed + anti-AI grain finalize. The on-screen UI, any faces in it, and the wordmark are ALWAYS real assets composited via PIL — never AI-rendered. The paid steps (plates, i2v clips, VO, music) are separate capabilities; this ships the config + PIPELINE + FREE assembly and the recipe orchestrates the spend. Use for the cgi-app-sizzle video format.
status: active

render-cgi-sizzle

Assembles the **3D-CGI app sizzle** video format: a phone floats in a CGI studio while five or six app features demo one per beat — each beat bursts REAL App Store UI elements out of the phone in 3D with rim-light + bokeh, then everything collapses back into the screen at a value-prop climax + a brand end card over a music bed. It reads as an Apple-keynote product film, not UGC and not a physical-product shoot.

The worked example (`scripts/config.example.json`) is the fictional trail-map app Trailvane: a gold phone in a smoky-black studio with an amaranth rim-light, an "Eryn" keynote-cadence VO, a premium-tech piano/synth bed, and an "Every trail worth walking. One app." climax. Those are **one example**, not defaults.

Choices

Creative calls the user makes (the recipe's `choices`, asked in one round before any paid step). The demo's value is only an example.

• **`studio_look`** — background, rim-light colour, bokeh; also the end-card canvas + bar.

Asked of the user; the demo used smoky-black `#0A0A0A` + amaranth `#E32652` rim-light.

• **`phone`** — the phone model/finish in every plate. Asked; the demo used a gold/champagne

titanium iPhone 15 Pro.

• **`narrator_voice`** — the VO voice. Asked; the demo used ElevenLabs "Eryn".
• **`vo_delivery`** — how the VO is written and read. Asked; the demo used a restrained

Apple-keynote spec-sheet cadence.

• **`music`** — the bed's genre/mood. Asked; the demo used a premium-tech underscore.

This capability ships the **recipe** (`scripts/config.example.json`) + a config→step map (`scripts/PIPELINE.md`) + the FREE-assembly how-to (`scripts/README.md`). It documents the FREE, deterministic assembly between the paid model calls:

• **nano-banana CGI plates** (paid, separate cap) render only the phone shell + the chosen

studio + rim-light + bokeh + placeholder burst shapes; the screen is left a blank warm glow on purpose, and every plate is `--anchor`ed on beat 1 so the phone/studio stay identical across beats.

• **PIL real-UI compositing (FREE)** — auto-detect the bright phone-screen bbox in each

plate and feather the REAL App Store screenshot into the bezel → `scene-NN-composite.png`; bake real burst-out overlays (e.g. climax portrait tiles, rim-light baked before rotation) around the plate. The on-screen UI + faces + wordmark are ALWAYS real assets — never AI-rendered, so no claim is invented and nothing reads fake.

• **Kling 3.0 i2v steady-float** (paid, separate cap) drives each composite; the burst-out

pops/settles while the phone + screen stay locked. Any beat Kling garbles drops to a **FREE Ken-Burns FFmpeg push-in** (`zoompan`, heavier on the climax) — the shipped demo used this path for the feature beats.

• **Assembly + finalize (FREE)** — dice + intercut concat (timeline locked from the measured

VO durations), audio mix (sidechain-duck the music under VO, loudnorm -14 LUFS master), PIL brand end card (real wordmark, never AI), then **1.15x speed + anti-AI grain** master.

See `scripts/README.md` for the full FREE-assembly detail and `scripts/PIPELINE.md` for the config-field → source-step map.

Run

Config-and-PIPELINE capability (no re-built runnable pipeline here). Copy the structure of `scripts/config.example.json` → `config.json`, fill the creative fields from the recipe choices (never the demo's values), rebuild the brand/beats/screens, and follow `scripts/PIPELINE.md`:

VO first (locks the timeline) → nano-banana CGI plates → PIL screen composites → burst-climax overlays → Kling i2v clips (Ken-Burns fallback per garbled beat) → PIL end card → captions → sidechain-ducked mix → 1.15x speed + grain.

Output: 1080x1920, ~22s H.264 (+ AAC music). 5-6 feature beats + PIL end card.

Contract

• **REAL UI, always PIL — never AI.** Every app screen, any face in it, and the wordmark is

the real asset composited via PIL. AI plates only ever render the phone shell, studio, bokeh, and placeholder burst shapes. This is the format's whole credibility and the guard against invented claims.

• **Kling for the float, Ken-Burns fallback per beat.** If a Kling beat garbles the burst-out

UI, distorts the phone, or animates the screen, fall that beat to a FREE Ken-Burns push-in — never ship a garbled beat.

• **Anchor every plate on beat 1** so the phone/studio read identical across beats (one shoot).
• **Timeline locked from the measured VO durations**, never planned word counts.
• The paid steps — nano-banana plates, Kling 3.0 i2v beats, ElevenLabs VO + music — are

separate capabilities (create-image-fal, create-video-fal, create-vo-elevenlabs, create-music-elevenlabs); the recipe orchestrates them and gates the spend.