Checklist the per-video bar — engine/sim
100.0/100
quantitative quality · weights learn from your reviews (
engine.sim.memory review longform-W29-Tue 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
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 6 scenes · 107.5s · comp_id + rendered still + tier + the script
| # | Layer (comp_id · still · tier) | Beat | Timecode | Motion | Logo | Audio | VO / on-screen / caption |
|---|---|---|---|---|---|---|---|
| 1 |  matches intent shared field signature-3d | open | 0–18.4s | spatial-parallax | icon·white-knockout | ♪ bed_in | The hub has a face, and that face is a website. Today, web design. You will learn how to have the agent build and ship a real page from a plain description, so your site stops being a thing you dread editing and becomes a thing you press run on. on-screen: A site that builds itself |
expected on screen: red ground · dodeca hero in the shared Signal Field · Mensor leads · node-graph · spatial-parallax · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual node-graphshape dodecaground redtreatment white-knockoutmotion spatial-parallaxpower summoninstrument summon→A site that builds itself | |||||||
| 2 |  matches intent NumberedList template | teach | 18.4–37.2s | kinetic-build | icon·white-knockout | ♪ node_lock | Write the page in words first. A headline, three sections, what each one says, who it is for. Save it as a markdown file. That file is your brief. The agent reads intent from words far better than it guesses from a blank canvas, so spend your time on the words. on-screen: Say the page out loud, in a |
expected on screen: red ground · a NumberedList panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id NumberedListvisual node-graphshape dodecaground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Say the page out loud, in a filecurate items, nodes | |||||||
| 3 |  matches intent ChecklistCard template | teach | 37.2–57.400000000000006s | kinetic-build | icon·liquid-chrome | ♪ node_lock | Hand the brief to Claude Code and ask for one HTML file with the styles inside it. It writes the markup, the spacing, the colors, the whole thing, then opens it so you can look. You point at what is ugly in plain language, it edits, you look again. Three rounds and it holds up. on-screen: Let it write the real HTML |
expected on screen: red ground · a ChecklistCard panel over a dimmed Signal Field · Mensor leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id ChecklistCardvisual node-graphshape dodecaground redtreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→Let it write the real HTMLcurate items, nodes | |||||||
| 4 |  matches intent BlueprintGrid template | teach | 57.4–75.4s | kinetic-build | icon·white-knockout | ♪ node_lock | Now put it on the internet. The agent calls a deploy tool through MCP and pushes the file to a host like Vercel. One command, a live link comes back. No dragging files, no dashboard maze. The same agent that wrote the page is the one that ships it. on-screen: Ship it with one deploy command |
expected on screen: red 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 redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Ship it with one deploy commandcurate nodes, rows | |||||||
| 5 |  matches intent StatScoreboard template | proof | 75.4–93.4s | receipts-count | icon·liquid-chrome | ♪ node_lock | The honest proof is the site you are headed to. The office pages were written, styled, and deployed by this exact loop, in the open. We did not hand it to a separate team and wait a month. The thing that proves the method is the method's own output. on-screen: This whole office site shipped this way |
expected on screen: red ground · a StatScoreboard panel over a dimmed Signal Field · Mensor leads · receipts · receipts-count · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id StatScoreboardvisual receiptsshape dodecaground redtreatment liquid-chromemotion receipts-countpower receiptsinstrument spotlight→This whole office site shipped this waycurate pillar, stats, statsLabels | |||||||
| 6 |  matches intent shared field signature-3d | resolve | 93.4–107.5s | coalescence | icon·white-knockout | ♪ bed_out | So write the brief, let the agent build the page, ship it with one command. Start with one page, your most important one. Then go see ours, built the same way, live at office dot temerarii dot xyz. on-screen: Build one page tonight at the office |
expected on screen: red ground · dodeca hero in the shared Signal Field · Mensor leads · coalescence · coalescence · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape dodecaground redtreatment white-knockoutmotion coalescencepower coalescenceinstrument coalescence→Build one page tonight at the office | |||||||
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)
youtubeBuild and Ship a Real Web Page From a Plain Description, Then Deploy It With One Command
This walkthrough covers web design done as a thing you press run on, not a thing you dread editing. Have an agent build and ship a real page from a plain description, start to finish.
The method:
- Say the page out loud, in a file. Write the page in words first, a headline, three sections, what each says, who it is for. Save it as a markdown file. That is your brief. The agent reads intent from words far better than it guesses from a blank canvas, so spend your time on the words.
- Let it write the real HTML. Hand the brief to a coding agent and ask for one HTML file with the styles inside. It writes the markup, spacing, colors, the whole thing, then opens it so you can look. You point at what is ugly in plain language, it edits, you look again. Three rounds and it holds up.
- Ship it with one deploy command. The agent calls a deploy tool through MCP and pushes the file to a host like Vercel. One command, a live link comes back. No dragging files, no dashboard maze. The same agent that wrote the page ships it.
The honest proof: the office pages were written, styled, and deployed by this exact loop, in the open. We did not hand it to a separate team.
Your next step: write the words for one page, let the agent build it, and deploy it. See it running at office.temerarii.xyz.
Keywords: web design, HTML, CSS, deploy, Vercel, MCP server, coding agent, build a website.