engine.sim.memory review longform-W40-Sun good|bad)| # | Layer (comp_id · still · tier) | Beat | Timecode | Motion | Logo | Audio | VO / on-screen / caption |
|---|---|---|---|---|---|---|---|
| 1 | ![]() matches intent shared field signature-3d | open | 0–37.0s | spatial-parallax | icon·liquid-chrome | ♪ bed_in | This week the calendar says one thing. Define who you are. Most brands skip that and jump straight to logos and ads. Today we do the boring part that everyone avoids, which is branding itself. Branding is not a color. It is the set of decisions about who you are for, what you stand against, and how you sound. By the end you will know how to pull those decisions out of your own head and write them down so a machine, or a new hire, can hold the line for you. We build The Big T-M this exact way, in public. on-screen: This week: define who you are |
expected on screen: red ground · tetra hero in the shared Signal Field · Nexus leads · node-graph · spatial-parallax · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual node-graphshape tetraground redtreatment liquid-chromemotion spatial-parallaxpower summoninstrument summon→This week: define who you are | |||||||
| 2 | ![]() first render · fix pending NumberedList templatecustom_element_dup | teach | 37.0–65.8s | kinetic-build | icon·white-knockout | ♪ node_lock | Step one. Open a plain text file. Not a deck, not a fancy tool, just a file. Call it brand-core. You are going to put every brand decision in here in plain words. We keep ours in the repo next to the code. The point is one place that everything else reads from. When the file is the source, your video, your site, and your emails can never quietly drift apart, because they all start from the same sentences. on-screen: Start with one source file |
expected on screen: red ground · a NumberedList panel over a dimmed Signal Field · Nexus leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id NumberedListvisual node-graphshape tetraground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Start with one source filecurate items, nodes | |||||||
| 3 | ![]() matches intent ChecklistCard template | teach | 65.8–96.4s | kinetic-build | icon·liquid-chrome | ♪ node_lock | Step two. Write one line that names the person you serve and the person you do not. Be rude about it. Ours says we are for owners who want to run the work, not watch a meeting about it. If your line could describe everyone, it describes no one. Paste that line into a model and ask it to argue the opposite side. If the model can defend the opposite easily, your line is too soft. Tighten it until the other side sounds dumb. on-screen: Name who it is for |
expected on screen: red ground · a ChecklistCard panel over a dimmed Signal Field · Nexus leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id ChecklistCardvisual node-graphshape tetraground redtreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→Name who it is forcurate items, nodes | |||||||
| 4 | ![]() matches intent JsonDiff template | teach | 96.4–124.10000000000001s | kinetic-build | icon·white-knockout | ♪ node_lock | Step three. List three things you are against. Hype with no method. Hidden pricing. Work you can never see. A brand gets its shape from what it refuses, the same way a knife gets its edge from what gets ground away. Write the three down in the file. Then ask a model to flag any sentence in your future copy that violates one of these. That turns a vibe into a rule a machine can check. on-screen: Pick your three enemies |
expected on screen: red ground · a JsonDiff panel over a dimmed Signal Field · Nexus leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id JsonDiffvisual node-graphshape tetraground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Pick your three enemiescurate fileName, lines, nodes | |||||||
| 5 | ![]() matches intent BlueprintGrid template | teach | 124.1–151.1s | kinetic-build | icon·liquid-chrome | ♪ node_lock | Step four. Voice is not a mood, it is a rule list. Write five rules a sixth-grader could follow. Short sentences. No empty hype words. Teach the real move. Dry humor, never goofy. No fake numbers. Save those rules in the file. Now feed any draft and the rule list to a model and ask, does this break a rule, and where. The model becomes a voice checker you never have to pay overtime. on-screen: Write the voice in five rules |
expected on screen: red ground · a BlueprintGrid panel over a dimmed Signal Field · Nexus leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id BlueprintGridvisual node-graphshape tetraground redtreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→Write the voice in five rulescurate nodes, rows | |||||||
| 6 | ![]() matches intent BuildLog template | teach | 151.1–179.1s | kinetic-build | icon·white-knockout | ♪ node_lock | Step five. Some rules are easy to enforce because they are just words. Take the empty hype words everyone reaches for, the ones that promise everything and teach nothing. Put them in a list inside your file. Then write a tiny check, ten lines of Python or a model prompt, that scans every script and flags any banned word before it ever ships. We run this on every line we write. It is unglamorous and it works. on-screen: Turn each rule into a banned list |
expected on screen: red ground · a BuildLog panel over a dimmed Signal Field · Nexus leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id BuildLogvisual node-graphshape tetraground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Turn each rule into a banned listcurate lines, nodes | |||||||
| 7 | ![]() matches intent DiffCard template | teach | 179.1–207.5s | kinetic-build | icon·liquid-chrome | ♪ node_lock | Step six. You do not have to write the brand alone. Open a coding agent like Claude Code in your terminal, hand it your half-filled file, and tell it to interview you. Ask it for ten sharp questions about your customer and your promise. Answer out loud, let it write your answers back into the file in clean sentences. You are still the author. The machine is just the patient note-taker that never gets tired of asking why. on-screen: Make the model interview you |
expected on screen: red ground · a DiffCard panel over a dimmed Signal Field · Nexus leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id DiffCardvisual node-graphshape tetraground redtreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→Make the model interview youcurate lines, nodes | |||||||
| 8 | ![]() matches intent StackTrace template | teach | 207.5–234.5s | kinetic-build | icon·white-knockout | ♪ node_lock | Step seven. The payoff comes from wiring. Point your site builder, your script writer, and your email tool at the same brand-core file through simple adapters called MCP servers. One file in, many surfaces out. Change your enemy list once, and the next render, the next article, and the next email all read the new version. The brand stays one thing because it physically comes from one place, not from everyone remembering the meeting. on-screen: Connect the file to everything |
expected on screen: red ground · a StackTrace panel over a dimmed Signal Field · Nexus leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id StackTracevisual node-graphshape tetraground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Connect the file to everythingcurate errMsg, errType, fix, frames, nodes | |||||||
| 9 | ![]() matches intent RankList template | proof | 234.5–262.2s | receipts-count | icon·liquid-chrome | ♪ node_lock | Here is the honest proof, no numbers attached. The whole calendar you are watching, every week, every day, was built from one source file run through these tools. The voice checker ran on this very script. The banned-word list flagged my drafts and I rewrote them. We did not hire a brand agency to tell us who we are. We opened the machine and made it hold our own line, and you can watch it live. on-screen: Watch us run on our own rules |
expected on screen: red ground · a RankList panel over a dimmed Signal Field · Nexus leads · receipts · receipts-count · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id RankListvisual receiptsshape tetraground redtreatment liquid-chromemotion receipts-countpower receiptsinstrument spotlight→Watch us run on our own rulescurate rows, statsLabels | |||||||
| 10 | ![]() matches intent shared field signature-3d | resolve | 262.2–290.2s | coalescence | icon·liquid-chrome | ♪ bed_out | So that is branding, the real version. One source file. A named customer. Three enemies. Five voice rules. A check that runs on every line. Do that and your brand becomes a thing a machine can defend, not a feeling that fades by Friday. The whole studio doing this in the open is live at office dot temerarii dot xyz. Go read our brand-core file and steal the structure. The next thing to build is your own. on-screen: See it at office.temerarii.xyz |
expected on screen: red ground · tetra hero in the shared Signal Field · Nexus leads · coalescence · coalescence · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape tetraground redtreatment liquid-chromemotion coalescencepower coalescenceinstrument coalescence→See it at office.temerarii.xyz | |||||||