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-W30-Sunkind longformweek W30date 2026-07-26campaign longform-youtubepillar performancebeat asset videoduration 260.1sground blackscenes 10

Checklist the per-video bar — engine/sim

98.6/100
plain languagevo coverageno dead airuniquenesscaption fitcompletenesscleanliness
quantitative quality · weights learn from your reviews (engine.sim.memory review longform-W30-Sun good|bad)
⚠ 1 flag(s) — not yet ship-ready: copy_generic · 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 10 scenes · 260.1s · comp_id + rendered still + tier + the script

#Layer (comp_id · still · tier)BeatTimecodeMotionLogoAudioVO / on-screen / caption
1s1
matches intent
shared field
signature-3d
open0–30.9sspatial-parallaxicon·wireframe♪ bed_in
This week the calendar says one thing: one idea, nine outputs. Most studios start a new project from zero every single time. We do the opposite. We take a single content plan and turn it into video, photos, a live stream, a script, and more, all from one source. Today we open the plan itself. By the end you will know how to write a content strategy that a machine can read and run, instead of a deck that sits in a folder and dies.
on-screen: One idea, nine outputs
expected on screen: black ground · dodeca hero in the shared Signal Field · Mensor leads · node-graph · spatial-parallax · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual node-graphshape dodecaground blacktreatment wireframemotion spatial-parallaxpower summoninstrument summon→One idea, nine outputs
2s2
matches intent
NumberedList
template
teach30.9–57.5skinetic-buildicon·wireframe♪ node_lock
Start with one file. Not a pretty document, a plain text file the model can read. We use a simple format called JSON, which is just labeled boxes. One box for the idea, one for the week, one for the audience. The whole point is that a human and a machine read the same file. No copy, no paste, no version six final final. One source, and everything else points back to it.
on-screen: Write one source file
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→Write one source filecurate items, nodes
3s3
matches intent
ChecklistCard
template
teach57.5–83.8skinetic-buildicon·wireframe♪ node_lock
Before you write a word of content, write down who it is for. We put the audience right in the file: who they are, what they already know, what they want next. Then we hand that to the model with a clear instruction. Read this audience, and do not write above their head. That one rule kills most bad marketing, because most bad marketing is written to impress, not to be understood.
on-screen: Name the audience first
expected on screen: black ground · a ChecklistCard panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id ChecklistCardvisual node-graphshape dodecaground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Name the audience firstcurate items, nodes
4s4
matches intent
SchematicCard
template
teach83.8–109.0skinetic-buildicon·wireframe♪ node_lock
Now the hard part. One idea, not nine. People think strategy means more ideas. It means fewer. We make the model argue with us. We say, here are five angles, pick the one a stranger would repeat to a friend, and tell me why the other four are weaker. When you force the cut, you find the idea that can carry an entire week instead of one tired post.
on-screen: Pick the one idea
expected on screen: black ground · a SchematicCard panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id SchematicCardvisual node-graphshape dodecaground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Pick the one ideacurate nodes
5s5
matches intent
CheatSheet
template
teach109.0–132.4skinetic-buildicon·wireframe♪ node_lock
One idea is still too big to film. So we break it into beats: the hook, the teach, the proof, the close. Each beat is one job. We store the beats in the same file, in order. The model never has to guess what comes next, because the order is written down. A beat list is the difference between a plan and a wish.
on-screen: Break it into beats
expected on screen: black ground · a CheatSheet panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id CheatSheetvisual node-graphshape dodecaground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Break it into beatscurate nodes, points, uses
6s6
matches intent
StackTrace
template
teach132.4–156.5skinetic-buildicon·wireframe♪ node_lock
Here is the move that gets you nine outputs. Take each beat and ask, what format does this beat want to live in. The hook wants a short vertical video. The proof wants a photo or a chart. The full teach wants the long form you are watching now. We write the format next to each beat, so the plan already knows what it will become.
on-screen: Map each beat to a format
expected on screen: black ground · a StackTrace panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id StackTracevisual node-graphshape dodecaground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Map each beat to a formatcurate errMsg, errType, fix, frames, nodes
7s7
matches intent
NodeGraphCard
template
teach156.5–181.3skinetic-buildicon·wireframe♪ node_lock
Now we let the model do the boring multiplication. One agent reads the beats and the formats, and writes a draft for each one: the caption, the script, the photo brief, the stream rundown. It is the same idea every time, said the right way for each place. You are not writing nine things. You are writing one thing and teaching a machine to translate it nine ways.
on-screen: Let the model fan it out
expected on screen: black ground · a NodeGraphCard panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id NodeGraphCardvisual node-graphshape dodecaground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Let the model fan it outcurate hub, nodes
8s8
matches intent
AnnotatedDiagram
template
teach181.3–205.8skinetic-buildicon·wireframe♪ node_lock
The trap with nine outputs is they drift apart and stop feeling like one campaign. So we keep a thread. Every output carries the same hook line and the same close. We tell the model to repeat the spine and vary the body. That way a person who sees the photo, then the video, then the stream, feels one voice, not three strangers wearing the same logo.
on-screen: Keep a thread between them
expected on screen: black ground · a AnnotatedDiagram panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id AnnotatedDiagramvisual node-graphshape dodecaground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Keep a thread between themcurate callouts, nodes
9s9
matches intent
ComparisonTable
template
proof205.8–232.10000000000002sreceipts-counticon·wireframe♪ node_lock
Does it work? Look at this week. One brief file became this video, the short clips, the photos, and the posts you are seeing, all from a single source. We did not hire a team of nine. We wrote one plan and let the machine fan it out, in public, with nothing hidden. That is the only proof we trust: we run it on ourselves before we ever sell it to you.
on-screen: We run this on us first
expected on screen: black ground · a ComparisonTable panel over a dimmed Signal Field · Mensor leads · receipts · receipts-count · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id ComparisonTablevisual receiptsshape dodecaground blacktreatment wireframemotion receipts-countpower receiptsinstrument spotlight→We run this on us firstcurate colA, colB, rows, statsLabels
10s10
matches intent
shared field
signature-3d
resolve232.1–260.1scoalescenceicon·wireframe♪ bed_out
So here is the takeaway. Content strategy is not a mood board, it is a file a machine can run. Write one source, name the audience, cut to one idea, break it into beats, map each beat to a format, then let the model fan it out while you guard the thread. You can watch the whole calendar, built from one file, live at office dot temerarii dot xyz. Go read the plan, then go write yours.
on-screen: See the whole plan live
expected on screen: black ground · dodeca hero in the shared Signal Field · Mensor leads · coalescence · coalescence · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape dodecaground blacktreatment wireframemotion coalescencepower coalescenceinstrument coalescence→See the whole plan live

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)

youtubeOne Idea, Nine Outputs: Turn a Single Content Plan Into Video, Photos, Posts, and More Most studios start a new project from zero every time. This walkthrough does the opposite: take a single content plan and turn it into video, photos, a live stream, a script, and more, all from one source file a machine can run. The method: - Write one source file. A plain text file the model can read, using JSON, which is just labeled boxes. One box for the idea, one for the week, one for the audience. - Name the audience first. Write down who it is for, what they already know, what they want next, right in the file. Hand that to the model: read this audience, write for them. - Pick the one idea. Strategy means fewer ideas, not more. Make the model argue with you. Here are five angles, pick the one a stranger would repeat to a friend, and say why the other four are weaker. - Break it into beats. The hook, the teach, the proof, the close. Each beat is one job, stored in the same file in order, so the model never has to guess what comes next. - Map each beat to a format. The hook wants a short vertical video, the proof wants a photo or a chart, the full teach wants the long form. This is the move that gets you nine outputs. - Let the model fan it out. One agent reads the beats and formats and writes a draft for each, the caption, the script, the photo brief, the stream rundown. - Keep a thread between them. Every output carries the same hook line and the same close, so the outputs do not drift apart and stop feeling like one campaign. This week, one brief file became this video, the short clips, the photos, and the posts you are seeing, from a single source. Content strategy is not a mood board, it is a file a machine can run. See the whole plan at office.temerarii.xyz. Keywords: content strategy, content repurposing, JSON brief, audience, beats, multi-platform content.

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