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-W35-Monkind longformweek W35date 2026-08-31campaign longform-youtubepillar performancebeat asset videoduration 157.2sground blackscenes 8

Checklist the per-video bar — engine/sim

93.4/100
plain languagevo coverageno dead airuniquenesscaption fitcompletenesscleanliness
quantitative quality · weights learn from your reviews (engine.sim.memory review longform-W35-Mon good|bad)
⚠ 4 flag(s) — not yet ship-ready: copy_genericgeneric_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 · 157.2s · comp_id + rendered still + tier + the script

#Layer (comp_id · still · tier)BeatTimecodeMotionLogoAudioVO / on-screen / caption
1s1
matches intent
shared field
signature-3d
open0–26.6sspatial-parallaxicon·wireframe♪ bed_in
The headline this week is about choosing an automation platform without copy-paste. Today we go one layer down, into software development, the part where you stop buying and start building. Most small teams think writing code is the scary option. It is often the calm one. We will show you when a few lines of your own code beat a tool you rent, and how to keep that code small enough to trust.
on-screen: When to build, not buy
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→When to build, not buy
2s2
matches intent
NumberedList
template
teach26.6–50.0skinetic-buildicon·ember-fill♪ node_lock
Here is the first move. Take the one service you depend on, read its API docs, and write a tiny wrapper: one function that takes plain inputs and returns plain outputs. Maybe twenty lines. Now your workflow calls your function, not the vendor directly. When the vendor changes, you fix one file. When you switch vendors, you swap one file. You own the seam.
on-screen: Wrap the API yourself
expected on screen: black ground · a NumberedList panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·ember-fill logo · caption bottom-left
spec (the prompt): comp_id NumberedListvisual node-graphshape dodecaground blacktreatment ember-fillmotion kinetic-buildpower morphinstrument morph+laser→Wrap the API yourselfcurate items, nodes
3s3
matches intent
ChecklistCard
template
teach50.0–75.9skinetic-buildicon·wireframe♪ node_lock
Never type a key into your code. Read it from an environment variable, a value the computer holds, not the file. In code that is one line that pulls the key from the room the program runs in. Now you can share the code, commit it, show it on a stream, and the key stays out. A platform that makes you paste keys into the workflow body is leaking by design.
on-screen: Read secrets from the room, not the
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→Read secrets from the room, not the filecurate items, nodes
4s4
matches intent
BlueprintGrid
template
teach75.9–100.4skinetic-buildicon·white-knockout♪ node_lock
You do not have to write all of it. Point a coding agent at your API docs and have it draft the wrapper, then you read every line before you trust it. The skill is not typing, it is reading code well enough to catch the one wrong line. Give the work away to the model, keep the judgment for yourself. That split is the whole job.
on-screen: Make the model write the boring part
expected on screen: black ground · a BlueprintGrid panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id BlueprintGridvisual node-graphshape dodecaground blacktreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Make the model write the boring partcurate nodes, rows
5s5
matches intent
LegacyCard
template
teach100.4–124.2skinetic-buildicon·liquid-chrome♪ node_lock
Our proof is that the studio's own tooling was written this way, small wrappers around real services, and it runs where anyone can look. We will not quote a speed-up number we cannot show you. What we can show is the running thing at office dot temerarii dot xyz, built from code you could read, not a black box we ask you to believe in.
on-screen: It builds itself in public
expected on screen: black ground · a LegacyCard panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left
spec (the prompt): comp_id LegacyCardvisual node-graphshape dodecaground blacktreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→It builds itself in publiccurate nodes
6s6
matches intent
BuildLog
template
teach124.2–147.2skinetic-buildicon·ember-fill♪ node_lock
So the takeaway: when a tool would lock you in, write the small wrapper instead and own that one seam. Read your keys from the environment. Let the model draft, you approve. Then go see software that builds in the open at office dot temerarii dot xyz. Start with one function. That is the next step, and it is smaller than you fear.
on-screen: Own one file at office.temerarii.xyz
expected on screen: black ground · a BuildLog panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·ember-fill logo · caption bottom-left
spec (the prompt): comp_id BuildLogvisual node-graphshape dodecaground blacktreatment ember-fillmotion kinetic-buildpower morphinstrument morph+laser→Own one file at office.temerarii.xyzcurate lines, nodes
7s7
first render · fix pending
StatScoreboard
templategeneric_scenecustom_element_dup
proof147.2–152.2sreceipts-counticon·wireframe♪ node_lock
Performance, demystified: tie every dollar to a result
on-screen: Chapter 6
expected on screen: black ground · a StatScoreboard panel over a dimmed Signal Field · Mensor leads · receipts · receipts-count · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id StatScoreboardvisual receiptsshape dodecaground blacktreatment wireframemotion receipts-countpower receiptsinstrument spotlight→Chapter 6curate pillar, stats, statsLabels
8s8
first render · fix pending
shared field
signature-3dgeneric_scene
resolve152.2–157.2scoalescenceicon·wireframe♪ bed_out
Performance, demystified: tie every dollar to a result
on-screen: Chapter 7
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→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)

youtubeWhen to build instead of buy: wrap the API yourself and keep the code small enough to trust This week is about choosing an automation platform without copy-paste. This goes one layer down into software development, where you stop buying and start building. Writing code feels like the scary option. It is often the calm one. This shows when a few lines of your own code beat a tool you rent. What you learn: - Take the one service you depend on, read its API docs, and write a tiny wrapper: one function that takes plain inputs and returns plain outputs, maybe twenty lines. Your workflow calls your function, not the vendor. When the vendor changes, you fix one file. - Never type a key into your code. Read it from an environment variable, a value the computer holds, not the file. Now you can share the code, commit it, and show it on a stream while the key stays out. A platform that makes you paste keys into the workflow body is leaking by design. - Point a coding agent at your API docs and have it draft the wrapper, then read every line before you trust it. The skill is reading code well enough to catch the one wrong line. Give the work to the model, keep the judgment. Our proof is that the studio's own tooling was written this way, small wrappers around real services, running where anyone can look. We will not quote a speed-up number we cannot show. What we can show is the running thing, built from code you could read, not a black box. See software that builds in the open at office.temerarii.xyz. Start with one function. Keywords: build vs buy, API wrapper, environment variables, coding agent, software development, vendor lock-in

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