Back Office · office.temerarii.xyz
One asset, all the way in — composition, the wireframe + storyboard, the output format stack, and the template, all read from the SAME content-index record. The expected output matches what /media surfaces for this post.
post social-W35-Mon-4kind threadweek W35date 2026-08-31campaign thread · Monpillar performancebeat Monasset videoduration 55.9sground blackscenes 9

Checklist the per-video bar — engine/sim

85.1/100
plain languagevo coverageno dead airuniquenesscaption fitcompletenesscleanliness
quantitative quality · weights learn from your reviews (engine.sim.memory review social-W35-Mon-4 good|bad)
⚠ 4 flag(s) — not yet ship-ready: copy_generictoo_complexdup_sequence_across_assetsdup_set_across_assets · see docs/strategy/VIDEO-CHECKLIST.md

Composition comp · template family · expected output

composition SceneReelfamily / template SignalFieldReel
9:16 Reelpending1:1 Squarepending16:9 Widepending9:16 4Kpending1:1 4Kpending16:9 4KpendingGIF (SMS)pending
render pending — silent master not yet on disk
expected output: 0/7 rendered — same matrix the /media preview surfaces for this asset.

Composition layer × scene 9 scenes · 55.9s · comp_id + rendered still + tier + the script

#Layer (comp_id · still · tier)BeatTimecodeMotionLogoAudioVO / on-screen / caption
1s1
matches intent
shared field
signature-3d
open0–5.9sspatial-parallaxicon·white-knockout♪ bed_in
Old copywriting meant a writer staring at a blank page, hoping the headline would arrive.
on-screen: Writers staring at a blank page
expected on screen: black ground · dodeca hero in the shared Signal Field · Mensor leads · node-graph · spatial-parallax · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual node-graphshape dodecaground blacktreatment white-knockoutmotion spatial-parallaxpower summoninstrument summon→Writers staring at a blank page
2s2
matches intent
shared field
signature-3d
hook5.9–12.9skinetic-buildicon·ember-fill♪ node_lock
Now we brief the model the way you brief a junior writer, with proof and a real reader.
on-screen: Brief the model like a writer
expected on screen: black ground · dodeca hero in the shared Signal Field · Mensor leads · mark · kinetic-build · icon·ember-fill logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual markshape dodecaground blacktreatment ember-fillmotion kinetic-buildpower laser-lockinstrument laser-trace→Brief the model like a writer
3s3
matches intent
NumberedList
template
teach12.9–18.8skinetic-buildicon·wireframe♪ node_lock
Feed it your real customer reviews, so the copy borrows the words buyers already use.
on-screen: Feed it your actual reviews
expected on screen: black ground · a NumberedList panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id NumberedListvisual node-graphshape dodecaground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Feed it your actual reviewscurate items, nodes
4s4
matches intent
ComparisonTable
template
proof18.8–25.1sreceipts-counticon·white-knockout♪ node_lock
Ask for twenty angles, not one finished ad, then a human picks and sharpens the best.
on-screen: Generate angles, not final lines
expected on screen: black ground · a ComparisonTable panel over a dimmed Signal Field · Mensor leads · receipts · receipts-count · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id ComparisonTablevisual receiptsshape dodecaground blacktreatment white-knockoutmotion receipts-countpower receiptsinstrument spotlight→Generate angles, not final linescurate colA, colB, rows, statsLabels
5s5
matches intent
CodeWindow
template
diff25.1–31.700000000000003scrossfade-8ficon·liquid-chrome♪ node_lock
The model is a draft engine, the edit is where taste and truth get put back in.
on-screen: The model drafts, you edit
expected on screen: black ground · a CodeWindow panel over a dimmed Signal Field · Mensor leads · code · crossfade-8f · icon·liquid-chrome logo · caption bottom-left
spec (the prompt): comp_id CodeWindowvisual codeshape dodecaground blacktreatment liquid-chromemotion crossfade-8fpower morphinstrument morph→The model drafts, you editcurate codeLines, windowTitle
6s6
matches intent
ChecklistCard
template
teach31.7–36.7skinetic-buildicon·ember-fill♪ node_lock
Run two clearly different openers against each other, not ten near-identical ones.
on-screen: Test two openers, not ten
expected on screen: black ground · a ChecklistCard panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·ember-fill logo · caption bottom-left
spec (the prompt): comp_id ChecklistCardvisual node-graphshape dodecaground blacktreatment ember-fillmotion kinetic-buildpower morphinstrument morph+laser→Test two openers, not tencurate items, nodes
7s7
matches intent
KpiGrid
template
proof36.7–43.0sreceipts-counticon·wireframe♪ node_lock
Whatever opener wins becomes a line in the next brief, so the model keeps getting sharper.
on-screen: The winner teaches the next brief
expected on screen: black ground · a KpiGrid panel over a dimmed Signal Field · Mensor leads · receipts · receipts-count · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id KpiGridvisual receiptsshape dodecaground blacktreatment wireframemotion receipts-countpower receiptsinstrument spotlight→The winner teaches the next briefcurate kpis, statsLabels
8s8
matches intent
NodeGraphCard
template
step43.0–49.3skinetic-buildicon·white-knockout♪ node_lock
Step one is pasting your reviews into the brief, before you ask for a single headline.
on-screen: Start the brief with reviews
expected on screen: black ground · a NodeGraphCard panel over a dimmed Signal Field · Mensor leads · pipeline · kinetic-build · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id NodeGraphCardvisual pipelineshape dodecaground blacktreatment white-knockoutmotion kinetic-buildpower throwinstrument laser-trace→Start the brief with reviewscurate hub, nodes, stages, steps
9s9
matches intent
shared field
signature-3d
resolve49.3–55.9scoalescenceicon·white-knockout♪ bed_out
Machine speed, human taste, the reader never feels the seam. That is copy at The Big T-M.
on-screen: Machine speed, human taste
expected on screen: black ground · dodeca hero in the shared Signal Field · Mensor leads · coalescence · coalescence · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape dodecaground blacktreatment white-knockoutmotion coalescencepower coalescenceinstrument coalescence→Machine speed, human taste

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)

tiktokAI copywriting isn't 'write me an ad.' Feed the model your real customer reviews so it borrows the words buyers use, ask for 20 angles not one finished line, then a human edits. Test two different openers, and the winner teaches the next brief. (AI-assisted edit.)
instagramBrief the model like a junior writer. Feed it your real reviews. Generate angles, not final lines. Human edits in taste and truth. Winner teaches the next brief. #copywriting #aiwriting #marketing #content #temerarii
linkedinTakeaway for anyone using AI to write copy: don't ask for the finished ad. Brief the model like a junior writer. Feed it your real customer reviews so the copy borrows the words buyers already use. Ask for twenty angles rather than one polished line, then have a human pick and sharpen the strongest. Test two clearly different openers against each other, not ten near-identical ones, and feed the winning line into the next brief so the model keeps getting sharper. Machine speed, human taste; the reader never feels the seam.
xCopywriting truth: don't ask AI for the finished ad. Feed it your real reviews so it uses buyers' words, ask for 20 angles, a human edits, test two different openers, and the winner teaches the next brief. temerarii.com
facebookOld copywriting meant staring at a blank page. The AI method: brief the model like a junior writer. Feed it your real customer reviews so the copy borrows the words buyers already use, ask for twenty angles instead of one finished line, then edit in taste and truth. Test two different openers and let the winner shape the next brief. Machine speed, human taste. More at temerarii.com.
threadsAI copywriting isn't 'write me an ad.' Brief the model like a junior writer: feed it your real reviews so it uses buyers' words, ask for 20 angles not one finished line, then a human edits. Test two different openers, and the winner teaches the next brief.
pinterestAI copywriting workflow: brief the model with real customer reviews, generate multiple angles instead of finished ads, edit for taste and truth, A/B test two distinct openers, and feed the winner into the next brief. Evergreen marketing copy and ad writing guide.
blueskyCopywriting truth: don't ask AI for the finished ad. Feed it your real reviews so it uses buyers' words, ask for 20 angles, a human edits, test two openers, and the winner teaches the next brief. temerarii.com
youtubeAI Copywriting Done Right: Reviews In, Angles Out, Human Edits Old copywriting meant a writer staring at a blank page. The method: brief the model like a junior writer. Feed it your real customer reviews so the copy borrows the words buyers already use. Ask for twenty angles instead of one finished ad, then have a human pick and sharpen the best. Test two clearly different openers, not ten near-identical ones, and feed the winning line into the next brief so the model keeps getting sharper. Machine speed, human taste. More at temerarii.com. #copywriting #aiwriting #marketing

Cross-links every lens is a view on this one record