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-W33-Sunkind longformweek W33date 2026-08-16campaign longform-youtubepillar it_devbeat asset videoduration 214.0sground blackscenes 10

Checklist the per-video bar — engine/sim

100.0/100
plain languagevo coverageno dead airuniquenesscaption fitcompletenesscleanliness
quantitative quality · weights learn from your reviews (engine.sim.memory review longform-W33-Sun good|bad)
✓ all static checks pass — one-focal/scene · tier-by-beat · one-track caption · colorway · cast+shape correct · no banned/fabricated. (audio + visual tiers verify on the rendered finals — Phase 2)

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 · 214.0s · comp_id + rendered still + tier + the script

#Layer (comp_id · still · tier)BeatTimecodeMotionLogoAudioVO / on-screen / caption
1s1
matches intent
shared field
signature-3d
open0–28.8sspatial-parallaxicon·white-knockout♪ bed_in
This week is about one simple idea. Tie every campaign you run to actual revenue, so you stop guessing and start seeing what works. Today we go deep on analytics, the part most people skip. By the end you will know how to pull your own numbers, connect them to sales, and tell which post or ad actually paid for itself. No dashboards full of pretty graphs that mean nothing. Just the path from a click to a dollar.
on-screen: Every campaign, traced to money
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→Every campaign, traced to money
2s2
matches intent
NumberedList
template
teach28.8–48.6skinetic-buildicon·wireframe♪ node_lock
Start by naming the moment money happens. A purchase. A booking. A signed quote. Pick one. In Google Analytics 4, open Admin, then Events, then mark that one event as a key event. That single switch tells the whole system which action you actually care about. Everything else is noise until you do this.
on-screen: Pick one event that means money
expected on screen: black ground · a NumberedList panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id NumberedListvisual node-graphshape boxground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Pick one event that means moneycurate items, nodes
3s3
matches intent
ChecklistCard
template
teach48.6–68.8skinetic-buildicon·wireframe♪ node_lock
Now make every link tell on itself. Add UTM tags to the end of your campaign links. Source, medium, and campaign name. So instead of one blurry pile of visitors, you get a clean label on each one. The man who tags his links knows where his money sleeps. The man who guesses just hopes.
on-screen: Tag the link, not your memory
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→Tag the link, not your memorycurate items, nodes
4s4
matches intent
SchematicCard
template
teach68.8–89.3skinetic-buildicon·wireframe♪ node_lock
Link your ad account to your analytics. In GA4, open Admin, then Product Links, and connect Google Ads. Now the cost of a click and the sale it caused live in the same table. You can finally divide one by the other. That number, money in over money spent, is the only ad score that matters.
on-screen: Connect ads to real sales
expected on screen: black ground · a SchematicCard panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id SchematicCardvisual node-graphshape boxground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Connect ads to real salescurate nodes
5s5
matches intent
AnnotatedDiagram
template
teach89.3–112.69999999999999skinetic-buildicon·wireframe♪ node_lock
Here is the move almost nobody makes. Wire an MCP server into your analytics, point a coding agent like Claude Code at it, and let the model pull your real numbers. Not made-up examples. Your traffic, your sales, your worst week. The model can read it and tell you in plain words which channel is quietly carrying you and which one is just expensive.
on-screen: Let a model read your own data
expected on screen: black ground · a AnnotatedDiagram panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id AnnotatedDiagramvisual node-graphshape boxground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Let a model read your own datacurate callouts, nodes
6s6
matches intent
JsonDiff
template
teach112.7–131.5skinetic-buildicon·wireframe♪ node_lock
Once the model holds your data, ask it the dull question. Which campaign brought money but few clicks. That campaign is your quiet winner, the one a vanity chart would hide. Cheap attention is a trap. Paid attention is a business. Tell the model to sort by revenue, never by likes.
on-screen: Ask the boring question first
expected on screen: black ground · a JsonDiff panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id JsonDiffvisual node-graphshape boxground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Ask the boring question firstcurate fileName, lines, nodes
7s7
matches intent
BuildLog
template
teach131.5–152.8skinetic-buildicon·wireframe♪ node_lock
Most tools give all the credit to the last click. That is lazy and it lies. In GA4, open the attribution settings and switch the model so earlier touches get credit too. The blog post that started the trust still counts. Now you stop killing the campaigns that warm people up and only look like they do nothing.
on-screen: Watch the path, not the last click
expected on screen: black ground · a BuildLog panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id BuildLogvisual node-graphshape boxground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Watch the path, not the last clickcurate lines, nodes
8s8
matches intent
BlueprintGrid
template
teach152.8–172.60000000000002skinetic-buildicon·wireframe♪ node_lock
Build a tiny weekly habit. Have the agent write one short report every Monday. Revenue by channel, cost by channel, and the gap between them. Save it to a file. Over a few weeks you get a record you can trust more than your gut. A gut has opinions. A logged file has receipts.
on-screen: Write the number down weekly
expected on screen: black ground · a BlueprintGrid panel over a dimmed Signal Field · Faber leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left
spec (the prompt): comp_id BlueprintGridvisual node-graphshape boxground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Write the number down weeklycurate nodes, rows
9s9
matches intent
StatScoreboard
template
proof172.6–191.7sreceipts-counticon·wireframe♪ node_lock
This is not a theory we sell from a stage. We run The Big T-M on the same wiring. The whole calendar you are watching was built from one source file, and the same agent reads what each piece earns. We point the camera at ourselves first, then show you the tape.
on-screen: We run our own books this way
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→We run our own books this waycurate pillar, stats, statsLabels
10s10
matches intent
shared field
signature-3d
resolve191.7–214.0scoalescenceicon·white-knockout♪ bed_out
So the takeaway is small and stubborn. Name the money event, tag your links, connect your ad spend, and let a model read your real data out loud. Do that and every campaign starts to answer one question. Did you pay for yourself. You can see the whole machine running in public at office.temerarii.xyz. Watch it, then go wire your own.
on-screen: See it at office.temerarii.xyz
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→See it at office.temerarii.xyz

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)

youtubeTie every campaign to revenue with GA4, UTM tags, and a model reading your real data This goes deep on analytics, the part most people skip. By the end you can pull your own numbers, connect them to sales, and tell which post or ad actually paid for itself. No dashboards full of pretty graphs that mean nothing, just the path from a click to a dollar. What you learn, step by step: - Name the moment money happens (a purchase, a booking, a signed quote) and mark it as a key event in GA4 under Admin, then Events. That switch tells the system which action you care about. - Add UTM tags to every campaign link, source, medium, and campaign name, so each visitor gets a clean label instead of one blurry pile. - Link Google Ads to GA4 under Admin, then Product Links, so the cost of a click and the sale it caused sit in the same table and you can divide one by the other. - Wire an MCP server into your analytics, point Claude Code at it, and let the model pull your real numbers and tell you in plain words which channel is quietly carrying you. - Ask the dull question: which campaign brought money but few clicks. That is your quiet winner a vanity chart would hide. - Switch the attribution model so earlier touches get credit, not just the last click, and have the agent write one short Monday report of revenue and cost by channel. We run our studio on the same wiring: the calendar was built from one source file and the same agent reads what each piece earns. We point the camera at ourselves first. See the whole machine running in public at office.temerarii.xyz, then go wire your own. Keywords: GA4 analytics, UTM tracking, attribution, Google Ads, MCP, revenue tracking

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