Checklist the per-video bar — engine/sim
76.1/100
quantitative quality · weights learn from your reviews (
engine.sim.memory review thread-minus-the-hype-W40-Thu good|bad)⚠ 7 flag(s) — not yet ship-ready: copy_genericlow_vo_coveragedead_airgeneric_scene · see docs/strategy/VIDEO-CHECKLIST.md
Composition comp · template family · expected output
9:16 Reelrendered1:1 Squarepending16:9 Widepending9:16 4Kpending1:1 4Kpending16:9 4KpendingGIF (SMS)pending
expected output: 1/7 rendered — same matrix the /media preview surfaces for this asset.
Composition layer × scene 5 scenes · 26.3s · comp_id + rendered still + tier + the script
| # | Layer (comp_id · still · tier) | Beat | Timecode | Motion | Logo | Audio | VO / on-screen / caption |
|---|---|---|---|---|---|---|---|
| 1 |  matches intent shared field signature-3d | hook | 0–5.4s | kinetic-build | icon·wireframe | ♪ node_lock | RAG gets sold as a trick: point a model at your docs and it answers anything. on-screen: "Just add a chatbot to your docs" |
expected on screen: white ground · torus hero in the shared Signal Field · Nuntius leads · mark · kinetic-build · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual markshape torusground whitetreatment wireframemotion kinetic-buildpower laser-lockinstrument laser-trace→"Just add a chatbot to your docs" | |||||||
| 2 |  first render · fix pending DiffCard templatedead_air | diff | 5.4–10.4s | crossfade-8f | icon·wireframe | ♪ node_lock | Here is the real work: you split the docs into small chunks, and the bad chunks decide the bad answers. on-screen: Reality: chunking is the whole game |
expected on screen: white ground · a DiffCard panel over a dimmed Signal Field · Nuntius leads · code · crossfade-8f · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id DiffCardvisual codeshape torusground whitetreatment wireframemotion crossfade-8fpower morphinstrument morph→Reality: chunking is the whole gamecurate codeLines, lines | |||||||
| 3 |  matches intent NumberedList template | teach | 10.4–16.3s | kinetic-build | icon·wireframe | ♪ node_lock | The real cost of lazy retrieval is a support bot that lies to your customers with total confidence. on-screen: The cost: confident wrong answers |
expected on screen: white ground · a NumberedList panel over a dimmed Signal Field · Nuntius leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id NumberedListvisual node-graphshape torusground whitetreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→The cost: confident wrong answerscurate items, nodes | |||||||
| 4 |  first render · fix pending ComparisonTable templatedead_airgeneric_scene | proof | 16.3–21.3s | receipts-count | icon·wireframe | ♪ node_lock | Minus the Hype |
expected on screen: white ground · a ComparisonTable panel over a dimmed Signal Field · Nuntius leads · receipts · receipts-count · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id ComparisonTablevisual receiptsshape torusground whitetreatment wireframemotion receipts-countpower receiptsinstrument spotlight→Minus the Hypecurate colA, colB, rows, statsLabels | |||||||
| 5 |  first render · fix pending shared field signature-3ddead_airgeneric_scene | resolve | 21.3–26.3s | coalescence | icon·wireframe | ♪ bed_out | Minus the Hype |
expected on screen: white ground · torus hero in the shared Signal Field · Nuntius leads · coalescence · coalescence · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape torusground whitetreatment wireframemotion coalescencepower coalescenceinstrument coalescence→Minus the Hype | |||||||
Format stack 3 aspects · same scenes[], re-cropped
9:16
1080×1920
Stories · TikTok · YouTube Shorts · Reels
1:1
1080×1080
LinkedIn · Facebook · Instagram
16:9
1920×1080
X/Twitter · YouTube · LinkedIn video
Channels 9 destinations
LinkedInX/TwitterYouTubeInstagramFacebookThreadsTikTokPinterestBluesky
Social captions supplemental published copy · per channel (comp_id level)
tiktokRAG gets sold as a trick: point a model at your docs and it answers anything. The real work is chunking. Split the docs into small pieces, because bad chunks decide the bad answers. Skip it and you ship a bot that lies with total confidence. (AI-assisted)
instagramRAG is sold as a trick: point a model at your docs, done.
The real work is chunking.
Split docs into small pieces. Bad chunks = bad answers.
Skip it and the bot lies, confidently.
#rag #ai #chatbots #llm #aiengineering
linkedinRAG gets sold as a simple trick: point a model at your documents and it answers anything. The real work is the part nobody demos.
Chunking is the whole game. You split the docs into small pieces, and the bad chunks decide the bad answers. Get retrieval lazy and you ship a support bot that lies to your customers with total confidence.
That is the honest cost. The method is here, do the chunking work.
xRAG is sold as a trick: point a model at your docs and it answers anything.
The real work is chunking. Split docs into small pieces, because bad chunks decide the bad answers.
Skip it and your bot lies with confidence.
facebookRAG gets sold as a simple trick: point a model at your documents and it answers anything. The real work is chunking. You split the docs into small pieces, and the bad chunks decide the bad answers. Skip that step and you ship a support bot that lies to your customers with total confidence. The method is the whole post. Do the boring part first.
threadsRAG is sold as a trick: point a model at your docs and it answers anything.
The real work is chunking. Split the docs into small pieces, because bad chunks decide the bad answers.
Get lazy with retrieval and you ship a bot that lies to customers, confidently.
pinterestWhy does your RAG chatbot give wrong answers? The honest reason for builders: chunking is the whole game. Split your documents into small, clean pieces, because bad chunks cause confident wrong answers. A practical retrieval-augmented generation tip for anyone adding AI to their docs.
blueskyRAG is sold as a trick: point a model at your docs and it answers anything. The real work is chunking. Split docs into small pieces, because bad chunks decide the bad answers. Skip it and the bot lies with confidence.
youtubeTitle: RAG, Minus the Hype: Chunking Is the Whole Game
RAG gets sold as a trick, point a model at your documents and it answers anything. The real work is chunking. We show why splitting your docs into small, clean pieces decides everything, and name the cost of skipping it: a support bot that lies to your customers with total confidence. The method is the whole video.