Checklist the per-video bar — engine/sim
100.0/100
quantitative quality · weights learn from your reviews (
engine.sim.memory review longform-W39-Mon 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 · 123.3s · 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 | 0–20.9s | spatial-parallax | icon·liquid-chrome | ♪ bed_in | The measurable stack, day two. Today is search. Most SEO advice is guessing about words you hope people type. We are going to stop guessing and use your own search data instead. By the end you will know how to pull the exact questions people already ask, and let a model rewrite your pages to answer them. on-screen: SEO you can actually measure |
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→SEO you can actually measure | |||||||
| 2 |  matches intent NumberedList template | teach | 20.9–41.8s | kinetic-build | icon·white-knockout | ♪ node_lock | First move. Open Google Search Console and connect it to your agent with a Search Console MCP server. Now pull your own query report, the real searches that already showed your site. These are not made up keywords. They are the actual words of people who almost found you. That is the only list worth working from. on-screen: Pull the queries you already get |
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→Pull the queries you already getcurate items, nodes | |||||||
| 3 |  matches intent ChecklistCard template | teach | 41.8–63.4s | kinetic-build | icon·liquid-chrome | ♪ node_lock | Second move. Sort that report for queries where you rank near the bottom of page one. Position eight, nine, ten. Those are the near misses. A small push there moves you into the spots people actually click. Ask the agent to list them, sorted by how often they get shown. You now have a to-do list written by reality. on-screen: Find the near misses |
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→Find the near missescurate items, nodes | |||||||
| 4 |  matches intent TerminalRun template | teach | 63.4–84.7s | kinetic-build | icon·white-knockout | ♪ node_lock | Third move. Hand the agent one near-miss query and the page that ranks for it. Tell it to rewrite the page so the question is answered in the first paragraph, plainly, with the real query in the heading. The model is not inventing a topic. It is matching your page to a question you can prove people ask. on-screen: Let the model rewrite from data |
expected on screen: red ground · a TerminalRun panel over a dimmed Signal Field · Nexus leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id TerminalRunvisual node-graphshape tetraground redtreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Let the model rewrite from datacurate nodes | |||||||
| 5 |  matches intent KpiGrid template | proof | 84.7–104.2s | receipts-count | icon·liquid-chrome | ♪ node_lock | Proof. We do this on ourselves, The Big T-M. The pages that pull search traffic for us are the ones where we gave the method away, like this one. We do not chase tricks. We answer the real question fully, and the search engine notices. No invented numbers here, just the honest pattern. on-screen: We rank for what we teach |
expected on screen: red ground · a KpiGrid panel over a dimmed Signal Field · Nexus leads · receipts · receipts-count · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id KpiGridvisual receiptsshape tetraground redtreatment liquid-chromemotion receipts-countpower receiptsinstrument spotlight→We rank for what we teachcurate kpis, statsLabels | |||||||
| 6 |  matches intent shared field signature-3d | resolve | 104.2–123.30000000000001s | coalescence | icon·liquid-chrome | ♪ bed_out | The takeaway. Search is a feedback loop, not a one-time chore. Pull your queries, fix your near misses, ship, and check Search Console again in two weeks to see what moved. Start with one page today. You can see how we run our own search loop at office dot temerarii dot xyz. on-screen: Ship one page, then check back |
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→Ship one page, then check back | |||||||
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)
youtubeStop guessing keywords and use your own search data instead
Most SEO advice is guessing about words you hope people type. This is how to use your own search data to find the exact questions people already ask, and let a model rewrite your pages to answer them.
The method:
- Pull the queries you already get. Open Google Search Console and connect it to your agent with a Search Console MCP server, then pull your own query report, the real searches that already showed your site. These are the actual words of people who almost found you.
- Find the near misses. Sort the report for queries where you rank near the bottom of page one, position eight, nine, ten. A small push there moves you into the spots people actually click. Sort by how often each gets shown.
- Let the model rewrite from data. Hand the agent one near-miss query and the page that ranks for it, and tell it to rewrite so the question is answered in the first paragraph, plainly, with the real query in the heading. It is matching your page to a question you can prove people ask.
The honest proof: we do this on ourselves. The pages that pull search traffic for us are the ones where we gave the method away, like this one. We answer the real question fully, and the search engine notices.
Pull your queries, fix your near misses, ship, and check Search Console again in two weeks. Start with one page today. See how we run our own search loop at office.temerarii.xyz.
Keywords: SEO, Google Search Console, near-miss queries, page two rankings, content optimization.