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 longform-W38-Monkind longformweek W38date 2026-09-21campaign longform-youtubepillar it_devbeat asset videoduration 139.7sground blackscenes 8

Checklist the per-video bar — engine/sim

95.1/100
plain languagevo coverageno dead airuniquenesscaption fitcompletenesscleanliness
quantitative quality · weights learn from your reviews (engine.sim.memory review longform-W38-Mon good|bad)
⚠ 3 flag(s) — not yet ship-ready: generic_scenecustom_element_dup · see docs/strategy/VIDEO-CHECKLIST.md

Composition comp · template family · expected output

composition LongFormChaptersfamily / template LongFormChapters
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 8 scenes · 139.7s · comp_id + rendered still + tier + the script

#Layer (comp_id · still · tier)BeatTimecodeMotionLogoAudioVO / on-screen / caption
1s1
matches intent
shared field
signature-3d
open0–21.6sspatial-parallaxicon·white-knockout♪ bed_in
Today is advertising, the part of performance marketing where it is easiest to set money on fire. The headline holds: stop guessing, start printing. By the end you will know how to run paid ads where the model watches the spend, finds the dead ones, and tells you exactly what to turn off. We are after sales, not applause.
on-screen: Make ads pay, not impress
expected on screen: black ground · box hero in the shared Signal Field · Faber leads · node-graph · spatial-parallax · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual node-graphshape boxground blacktreatment white-knockoutmotion spatial-parallaxpower summoninstrument summon→Make ads pay, not impress
2s2
matches intent
NumberedList
template
teach21.6–43.900000000000006skinetic-buildicon·ember-fill♪ node_lock
First, tell the ad platform what counts. In Meta or Google Ads, set the conversion to the real action, a purchase or a booking, and pass it back with their pixel or the conversions API. If the platform only knows about clicks, it will buy you clicks. Give it the thing that pays you, and it starts hunting for that instead.
on-screen: One goal the platform can see
expected on screen: black ground · a NumberedList panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·ember-fill logo · caption bottom-left
spec (the prompt): comp_id NumberedListvisual node-graphshape boxground blacktreatment ember-fillmotion kinetic-buildpower morphinstrument morph+laser→One goal the platform can seecurate items, nodes
3s3
matches intent
ChecklistCard
template
teach43.9–66.2skinetic-buildicon·wireframe♪ node_lock
Now wire a model into the ad account through the Google Ads or Meta marketing API. Ask it to pull, per campaign, what you spent and what came back. Put those two numbers side by side. Most owners never see them in the same row. The agent does it in one query, and the losers stop hiding inside a tidy average.
on-screen: Pull spend next to results
expected on screen: black ground · a ChecklistCard panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id ChecklistCardvisual node-graphshape boxground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Pull spend next to resultscurate items, nodes
4s4
matches intent
LogStream
template
teach66.2–87.1skinetic-buildicon·white-knockout♪ node_lock
Tell the model your rule in plain words: anything spending real money with no sales for a week, flag it to pause. Anything paying back well, flag it to feed more budget. The agent reads the account, applies your rule, and hands you a short list of turn-this-off, turn-this-up. You make the call. It did the digging.
on-screen: Cut the bottom, feed the top
expected on screen: black ground · a LogStream panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id LogStreamvisual node-graphshape boxground blacktreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Cut the bottom, feed the topcurate nodes, rows
5s5
matches intent
SchematicCard
template
teach87.1–108.39999999999999skinetic-buildicon·liquid-chrome♪ node_lock
We hold ourselves to the same line. When the studio promotes its own work, the agent checks the spend against the real result before we add a dollar, and it has told us to stop more than once. We listen. You can see how the studio operates, out loud, at office.temerarii.xyz, including the parts that did not work.
on-screen: We run our own spend this way
expected on screen: black ground · a SchematicCard panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left
spec (the prompt): comp_id SchematicCardvisual node-graphshape boxground blacktreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→We run our own spend this waycurate nodes
6s6
matches intent
ProcessFlow
template
teach108.4–129.70000000000002skinetic-buildicon·ember-fill♪ node_lock
So the takeaway: give the platform the action that is money, pull spend and return into one view, and let a model flag the dead ads on a rule you wrote. Do it tonight and you will likely find something to pause. The whole approach is running live at office.temerarii.xyz. Turn off one waster this week and watch.
on-screen: Pause the losers tonight
expected on screen: black ground · a ProcessFlow panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·ember-fill logo · caption bottom-left
spec (the prompt): comp_id ProcessFlowvisual node-graphshape boxground blacktreatment ember-fillmotion kinetic-buildpower morphinstrument morph+laser→Pause the losers tonightcurate nodes, steps
7s7
first render · fix pending
StatScoreboard
templategeneric_scenecustom_element_dup
proof129.7–134.7sreceipts-counticon·wireframe♪ node_lock
Custom apps + dashboards: own the code and the IP
on-screen: Chapter 6
expected on screen: black ground · a StatScoreboard panel over a dimmed Signal Field · Faber leads · receipts · receipts-count · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id StatScoreboardvisual receiptsshape boxground blacktreatment wireframemotion receipts-countpower receiptsinstrument spotlight→Chapter 6curate pillar, stats, statsLabels
8s8
first render · fix pending
shared field
signature-3dgeneric_scene
resolve134.7–139.7scoalescenceicon·white-knockout♪ bed_out
Custom apps + dashboards: own the code and the IP
on-screen: Chapter 7
expected on screen: black ground · box hero in the shared Signal Field · Faber leads · coalescence · coalescence · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape boxground blacktreatment white-knockoutmotion coalescencepower coalescenceinstrument coalescence→Chapter 7

Format stack 1 aspects · same scenes[], re-cropped

16:9
1920×1080
X/Twitter · YouTube · LinkedIn video

Channels 2 destinations

YouTubeBlog

Social captions supplemental published copy · per channel (comp_id level)

youtubeRun paid ads where a model watches the spend and tells you what to turn off Ads are the easiest place to set money on fire. This is how to run paid ads where the model pulls the spend, finds the dead ones, and hands you a turn-off, turn-up list. We are after sales, not applause. The method: - Give the platform one goal it can see. In Meta or Google Ads, set the conversion to the real action, a purchase or a booking, and pass it back with their pixel or conversions API. If the platform only knows clicks, it buys you clicks. - Pull spend next to results. Wire a model into the ad account through the Google Ads or Meta marketing API and ask it for, per campaign, what you spent and what came back. Most owners never see those two numbers in the same row. - Cut the bottom, feed the top. Tell the model your rule in plain words: anything spending real money with no sales for a week, flag to pause; anything paying back well, flag to feed. It reads the account, applies your rule, and hands you a short list. You make the call. The honest proof: when the studio promotes its own work, the agent checks spend against the real result before we add a dollar, and it has told us to stop more than once. We listen. Give the platform the action that is money, pull spend and return into one view, and let a model flag the dead ads on a rule you wrote. Turn off one waster this week. The approach runs live at office.temerarii.xyz. Keywords: paid advertising, ad spend tracking, Google Ads API, Meta marketing API, conversion tracking.

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