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-W37-Sunkind longformweek W37date 2026-09-13campaign longform-youtubepillar brandbeat asset videoduration 229.2sground redscenes 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-W37-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 · 229.2s · comp_id + rendered still + tier + the script

#Layer (comp_id · still · tier)BeatTimecodeMotionLogoAudioVO / on-screen / caption
1s1
matches intent
shared field
signature-3d
open0–28.8sspatial-parallaxicon·white-knockout♪ bed_in
This week the calendar says one thing: turn social listening into real decisions. Most people watch their mentions, feel something, and do nothing. We are going to do the opposite. We are going to take the raw chatter about a brand and turn it into a short list of moves you can actually make this week. No mood, no guessing. Listening is not a vibe. It is a pipeline, and a model can run most of it for you.
on-screen: We listen, then we decide
expected on screen: red ground · icosa hero in the shared Signal Field · Signum leads · node-graph · spatial-parallax · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual node-graphshape icosaground redtreatment white-knockoutmotion spatial-parallaxpower summoninstrument summon→We listen, then we decide
2s2
matches intent
NumberedList
template
teach28.8–54.0skinetic-buildicon·white-knockout♪ node_lock
Step one. Stop reading platforms one at a time. Use an Apify scraper to pull posts and comments about your name, your competitors, and your topic into one file. Point it at Reddit, at X, at the comment sections you care about. You end up with a plain spreadsheet of who said what and where. That file is the whole job. Everything after this is just reading it well.
on-screen: Pull the chatter into one place
expected on screen: red ground · a NumberedList panel over a dimmed Signal Field · Signum leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id NumberedListvisual node-graphshape icosaground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Pull the chatter into one placecurate items, nodes
3s3
matches intent
ChecklistCard
template
teach54.0–76.7skinetic-buildicon·liquid-chrome♪ node_lock
Step two. Feed that file to Claude and ask it to group the posts by what people actually want. Not by hashtag. By intent. People asking for help. People angry. People comparing you to someone else. People ready to buy. The model reads a thousand comments faster than you read ten, and it never gets bored or defensive about the mean ones.
on-screen: Let the model sort the noise
expected on screen: red ground · a ChecklistCard panel over a dimmed Signal Field · Signum leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left
spec (the prompt): comp_id ChecklistCardvisual node-graphshape icosaground redtreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→Let the model sort the noisecurate items, nodes
4s4
matches intent
JsonDiff
template
teach76.7–98.30000000000001skinetic-buildicon·white-knockout♪ node_lock
Step three. Ask the model for the three needs that show up the most, with real quotes attached. Quotes matter. A summary that says people want faster support is weak. A summary that pastes four people saying the same complaint in their own words is a decision waiting to happen. You are looking for the sentence that keeps repeating.
on-screen: Name the three loudest needs
expected on screen: red ground · a JsonDiff panel over a dimmed Signal Field · Signum leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id JsonDiffvisual node-graphshape icosaground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Name the three loudest needscurate fileName, lines, nodes
5s5
matches intent
SchematicCard
template
teach98.3–120.3skinetic-buildicon·liquid-chrome♪ node_lock
Step four, and this is the public relations move. In that same file, ask the model to flag the accounts people keep citing. The reporter they quote. The creator they tag. The forum everyone links to. That list is your earned-media map. You do not pitch everyone. You pitch the few voices your own audience already pulls into the room.
on-screen: Find who they already trust
expected on screen: red ground · a SchematicCard panel over a dimmed Signal Field · Signum leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left
spec (the prompt): comp_id SchematicCardvisual node-graphshape icosaground redtreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→Find who they already trustcurate nodes
6s6
matches intent
DiffCard
template
teach120.3–141.6skinetic-buildicon·white-knockout♪ node_lock
Step five, the partnership move. Look for tools, brands, and people that show up next to yours, where nobody is fighting. If your customers keep mentioning another product they use alongside yours, that is not a rival. That is a partner who already shares your audience. Have the model list every co-mention, ranked by how often it appears.
on-screen: Spot the partner hiding in comments
expected on screen: red ground · a DiffCard panel over a dimmed Signal Field · Signum leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id DiffCardvisual node-graphshape icosaground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Spot the partner hiding in commentscurate lines, nodes
7s7
matches intent
StackTrace
template
teach141.6–162.9skinetic-buildicon·liquid-chrome♪ node_lock
Step six, the event move. Search the file for time and place words. Meetups people wish existed. Questions that keep going unanswered. A city that keeps coming up. Listening tells you what room to build before you spend a dollar on the room. Ask the model: if these people were going to gather, what would they gather around.
on-screen: Read the room for an event
expected on screen: red ground · a StackTrace panel over a dimmed Signal Field · Signum leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left
spec (the prompt): comp_id StackTracevisual node-graphshape icosaground redtreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→Read the room for an eventcurate errMsg, errType, fix, frames, nodes
8s8
matches intent
NodeGraphCard
template
teach162.9–183.8skinetic-buildicon·white-knockout♪ node_lock
Step seven. Take your needs, your voices, your partners, and your event signal, and write one concrete action for each. One pitch. One outreach note. One small gathering. Ask the model to draft all four, in plain words, ready to send. A decision you cannot act on by Friday is not a decision. It is a note.
on-screen: Turn each finding into one action
expected on screen: red ground · a NodeGraphCard panel over a dimmed Signal Field · Signum leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id NodeGraphCardvisual node-graphshape icosaground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Turn each finding into one actioncurate hub, nodes
9s9
matches intent
StatScoreboard
template
proof183.8–205.8sreceipts-counticon·liquid-chrome♪ node_lock
Here is the honest part. We run this exact pipeline on ourselves. The whole campaign calendar you are watching came out of one source file, read by a model, the same way we just read those comments. We are not showing you a trick we keep for clients. We are showing you the actual room while the lights are on.
on-screen: This whole flow is public
expected on screen: red ground · a StatScoreboard panel over a dimmed Signal Field · Signum leads · receipts · receipts-count · icon·liquid-chrome logo · caption bottom-left
spec (the prompt): comp_id StatScoreboardvisual receiptsshape icosaground redtreatment liquid-chromemotion receipts-countpower receiptsinstrument spotlight→This whole flow is publiccurate pillar, stats, statsLabels
10s10
matches intent
shared field
signature-3d
resolve205.8–229.20000000000002scoalescenceicon·white-knockout♪ bed_out
So the takeaway is simple. Listening only counts when it ends in a move. Pull the chatter, group it by need, name the voices, the partners, and the room, then write the four actions. Your next step is small: open one scraper, point it at your own name, and read what comes back. The full thing is live at office dot temerarii dot xyz.
on-screen: See it at office.temerarii.xyz
expected on screen: red ground · icosa hero in the shared Signal Field · Signum leads · coalescence · coalescence · icon·white-knockout logo · caption bottom-left
spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape icosaground redtreatment 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)

youtubeTurn social listening into a short list of moves you can make this week Most people watch their mentions, feel something, and do nothing. We do the opposite: take the raw chatter about a brand and turn it into decisions you can act on. Listening is not a vibe, it is a pipeline, and a model runs most of it. The method: - Pull the chatter into one place. Use an Apify scraper to gather posts and comments about your name, your competitors, and your topic into one file. That file is the whole job. - Let the model sort by intent, not hashtag: people asking for help, people angry, people comparing you, people ready to buy. It reads a thousand comments faster than you read ten. - Name the three loudest needs, with real quotes attached. A summary that pastes four people saying the same complaint is a decision waiting to happen. - Find who they already trust. Flag the accounts people keep citing, the reporter, the creator, the forum. That is your earned-media map. You pitch the few voices your audience already pulls in. - Spot the partner hiding in the comments. List every brand mentioned next to yours, where nobody is fighting. A product people use alongside yours is a partner, not a rival. - Read the room for an event. Search for time and place words, meetups people wish existed, questions going unanswered. - Turn each finding into one concrete action: one pitch, one outreach note, one small gathering, drafted in plain words. The honest proof: we run this exact pipeline on ourselves. The whole calendar you are watching came out of one source file, read by a model, the same way. We are showing the actual room with the lights on. Open one scraper, point it at your own name, and read what comes back. The full thing is at office.temerarii.xyz. Keywords: social listening, audience intent, earned media, partner discovery, event planning.

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