Teardown — South America's Smallest Cat Weighs Two Kilos: 12 cuts, film + clips + every prompt
12 cuts across 2 motion tiers, 67% on tier 1. A finished episode, published, with every generated clip (12), every source still (12), and the exact still and motion prompt behind each cut.
Every AI video tutorial shows you the prompt that worked. This shows you all 12, including the dull ones.
One published wildlife documentary — a finished episode — opened up. The finished film, all 12 generated clips, the 12 stills they were animated from, and for each of the 12 cuts the exact still prompt and motion prompt that produced it, beside the engine tier it was ordered on.
Watch the published cut before you buy: https://youtu.be/R4IZ81rLTew
Why this genre is worth taking apart
Animals are where identity drift shows first. A prompt that names the species loosely gives you a hyena in cut nine.
Every still prompt here carries a species lock written as a positive count — ear shape, coat pattern, toe count — never as a list of things the animal must not be.
What is in the package
- The finished film, as published
- 12 generated clips — the raw shots before the edit
- 12 source stills — the frames each clip was animated from
- PROMPTS.md — one block per cut: framing, tier, still prompt, motion prompt
- production.json — the same record, machine-readable
The routing, in this episode
Cut split: 8 on tier 1, 4 on tier 2. Tier is assigned before a cut is ordered, and it decides the engine. A static shot on the expensive engine is pure waste; a hard-motion shot on the cheap one costs more, because the re-order costs more than the correct order would have.
What the episode argues
코드코드는 손바닥만 한 아메리카 최소 고양이인데, 야생 고양잇과 중 흑화(멜라니즘) 비율이 가장 높아 셋 중 하나가 흑표범처럼 새까맣고, 숲이 조각날수록 검은 개체가 더 늘어난다.
Three things you can read straight off the prompts
- Identity is locked with a positive count, never a negation — a forbidden thing tends to get summoned.
- Framing is stated as what fills the frame, not as a shot-size name. Shot-size vocabulary is interpreted loosely by every engine we have measured.
- Subject, action and environment are separate lines in the motion prompt. Collapsing them into one sentence is where drift starts.
Honest notice
- The narration and on-screen text are Korean. Every prompt and all documentation is in English — the prompts are the transferable part.
- Clips are re-encoded for download, not delivered at master bitrate. They are reference and working copies.
- You may use these files in your own commercial and personal work, including client work, and modify them freely. You may not resell or redistribute the package itself, in whole or in part, as a product — including uploading the clips to a stock library.
- This is a record of what one episode took, not a guarantee of a result. Generation models change behaviour between versions, sometimes reversing it, so treat the routing as a procedure you can re-measure rather than a fixed ranking.
—
Where this sits
- AI Video Prompt Craft — $29 — this episode is one worked example; the pack is the reusable standard behind it
- The Prompt Atlas — $49 — every prompt from 83 published episodes, not just this one
- 7 Ways AI Video Burns Your Credits — free — the seven measured credit leaks, at no cost
- The whole catalogue — teardowns, references, skill packs and free calculators, with what each one is for