From AI Prompt to Editable Canvas
A Framer Agents announcement that turns a typed task into a luminous canvas walkthrough and a finished site before a spare URL end card.
Field notes / why it works
A concrete request types out by 3.6 seconds, and the isolated blue prompt opens into the full canvas around 4.0 seconds, the strongest measured replay point. The long middle shot keeps changing its proof: agent controls at 10.5 seconds, code and interaction at 16–18 seconds, and a finished page at 21.5 seconds. A black logo and URL hold from roughly 30–35.6 seconds, making the destination easy to retain.
Plate 01 / The format in one breath
A polished product announcement starts with a specific job being typed into a glowing AI prompt. The film pulls back to show the prompt living inside a full design canvas, then moves through three concrete capabilities and a finished result before a quiet brand sign-off. The spoken promise is human control over an AI-assisted workflow; the screen supplies the proof.
Why it works
- The first 3.6 seconds turn an empty prompt into a complete, specific request, so viewers watch a small task resolve before the first cut at 6.84 seconds.
- The 4.0-second reveal converts an isolated glowing chat box into a working canvas; this is the strongest measured replay moment, 3.6–4.0 seconds.
- The proof escalates from agent controls at 10.5 seconds to code at 16.0, generative interaction at 17.8, and a composed page at 21.5, then gives the brand and URL the last five seconds.
Plate 02 / Format card
| Platform · aspect · length | YouTube · 16:9 · 640×360 source · 29.97 fps · 35.6 s |
| Pace | 3 detected shots, 2 hard cuts, 3.4 cuts/min; 3 measured step changes, 8.4 detected picture changes/min; mean shot 11.88 s. Continuous screen animation creates more visual activity than these cut counts imply. |
| Script | 76 spoken words; 147 wpm overall, 166 wpm while speaking; 6 measured sentences, about 13 words each; no spoken question. |
| Voice | One low-register, energetic, composed product narrator; no on-camera speaker. |
| On screen | Roughly 85% dark UI and interface motion, 15% minimal brand ending (inferred from the sheets); no live-action presenter. |
| Captions | No editorial subtitles. White UI labels and typed prompt text carry the visible words. |
| Sound | Continuous tonal bed (measured); electronic/ambient character is inferred. Exact instruments and cut effects are not established by the sheets. |
| Structure | Problem → agent/canvas promise → three feature proofs → finished page → CTA and black end card. |
Pro / 13 plates + recipe
The first plates are open. The full dissection is Pro.
Get the timed shots, type and sound specs, and the runnable recipe.
Open the full blueprint- Plate 03 / The hook, frame by frame (first 3.6 seconds)Pro
- Plate 04 / Structure (beat sheet)Pro
- Plate 05 / Shot grammarPro
- Plate 06 / Visual style guidePro
- Plate 07 / Voice and deliveryPro
- Plate 08 / Sound designPro
- Plate 09 / PackagingPro
- Plate 10 / Script templatePro
- Plate 11 / Production planPro
- Plate 12 / PromptsPro
- Plate 13 / Edit recipePro
- Plate 14 / QA checklistPro
- Plate 15 / Make it yoursPro
- Recipe / runnable promptPro