Checklist the per-video bar — engine/sim
98.3/100
quantitative quality · weights learn from your reviews (
engine.sim.memory review longform-W35-Thu good|bad)⚠ 1 flag(s) — not yet ship-ready: generic_scene · see docs/strategy/VIDEO-CHECKLIST.md
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 8 scenes · 163.2s · 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 | 0–24.5s | spatial-parallax | icon·ember-fill | ♪ bed_in | The week is about choosing an automation platform with no copy-paste. Today we point it at business intelligence: turning the data your tools already hold into answers you can act on. Most owners run on a gut feeling because the numbers live in five places. We will show you how to pull them into one view a model can read and explain, in plain language, on demand. on-screen: See your numbers, not a vibe |
expected on screen: black ground · torus hero in the shared Signal Field · Nuntius leads · node-graph · spatial-parallax · icon·ember-fill logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual node-graphshape torusground blacktreatment ember-fillmotion spatial-parallaxpower summoninstrument summon→See your numbers, not a vibe | |||||||
| 2 |  matches intent NumberedList template | teach | 24.5–46.1s | kinetic-build | icon·ember-fill | ♪ node_lock | Here is the core move. Wire an MCP server into a data source you own, like your analytics or your sales sheet, and let the model pull the numbers directly. No exporting, no copy-paste into a slide. The agent queries the live data, then writes the summary from what it actually found. You stop reformatting and start asking questions. on-screen: Wire a tool into your own data |
expected on screen: black ground · a NumberedList panel over a dimmed Signal Field · Nuntius leads · node-graph · kinetic-build · icon·ember-fill logo · caption bottom-left spec (the prompt): comp_id NumberedListvisual node-graphshape torusground blacktreatment ember-fillmotion kinetic-buildpower morphinstrument morph+laser→Wire a tool into your own datacurate items, nodes | |||||||
| 3 |  matches intent ChecklistCard template | teach | 46.1–67.4s | kinetic-build | icon·wireframe | ♪ node_lock | Before you trust a dashboard, define each number once, in writing. What counts as a lead, what counts as a sale, what date range. Put that definition next to the report. Half of all bad dashboards are not wrong math, they are two people meaning different things by the same word. Pin the meaning, and the arguments stop. on-screen: Define the number once |
expected on screen: black ground · a ChecklistCard panel over a dimmed Signal Field · Nuntius leads · node-graph · kinetic-build · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id ChecklistCardvisual node-graphshape torusground blacktreatment wireframemotion kinetic-buildpower morphinstrument morph+laser→Define the number oncecurate items, nodes | |||||||
| 4 |  matches intent AnnotatedDiagram template | teach | 67.4–92.2s | kinetic-build | icon·white-knockout | ♪ node_lock | When the agent gives you a number, make it show the query and the rows it used. A number with no source is a rumor. Ask the model to print the exact filter it ran, so you can check it. This is how you keep a confident-sounding answer honest. The point of business intelligence is not a pretty chart, it is a number you would defend out loud. on-screen: Make the model show its work |
expected on screen: black ground · a AnnotatedDiagram panel over a dimmed Signal Field · Nuntius leads · node-graph · kinetic-build · icon·white-knockout logo · caption bottom-left spec (the prompt): comp_id AnnotatedDiagramvisual node-graphshape torusground blacktreatment white-knockoutmotion kinetic-buildpower morphinstrument morph+laser→Make the model show its workcurate callouts, nodes | |||||||
| 5 |  matches intent LogStream template | teach | 92.2–116.0s | kinetic-build | icon·liquid-chrome | ♪ node_lock | One hard rule, for the model and for you: if the data is not there, the answer is I do not know, not a guessed figure. Teach the agent to say the data is missing rather than fill the gap with a number that sounds right. A made-up metric is worse than a blank, because someone will act on it. Honest gaps beat confident fiction. on-screen: Refuse the invented figure |
expected on screen: black ground · a LogStream panel over a dimmed Signal Field · Nuntius leads · node-graph · kinetic-build · icon·liquid-chrome logo · caption bottom-left spec (the prompt): comp_id LogStreamvisual node-graphshape torusground blacktreatment liquid-chromemotion kinetic-buildpower morphinstrument morph+laser→Refuse the invented figurecurate nodes, rows | |||||||
| 6 |  matches intent FlowSchematic template | teach | 116.0–137.3s | kinetic-build | icon·ember-fill | ♪ node_lock | Pull it together and you get one habit: one source of truth, queried live, with every number defined and sourced. Ask it anything, get an answer you can trace. We run our own calendar from a single source file for exactly this reason, so the answer is never spread across ten tools we have to reconcile by hand. on-screen: One source, many questions |
expected on screen: black ground · a FlowSchematic panel over a dimmed Signal Field · Nuntius leads · node-graph · kinetic-build · icon·ember-fill logo · caption bottom-left spec (the prompt): comp_id FlowSchematicvisual node-graphshape torusground blacktreatment ember-fillmotion kinetic-buildpower morphinstrument morph+laser→One source, many questionscurate nodes, stages | |||||||
| 7 |  matches intent StatScoreboard template | proof | 137.3–158.20000000000002s | receipts-count | icon·wireframe | ♪ node_lock | So your next step: pick the one number you check most, define it in a sentence, and wire a tool straight into the data behind it. Make the model show its work. Then see how we run on one source at office dot temerarii dot xyz. Stop guessing in private. Start asking your own data, out loud. on-screen: Ask your data at office.temerarii.xyz |
expected on screen: black ground · a StatScoreboard panel over a dimmed Signal Field · Nuntius leads · receipts · receipts-count · icon·wireframe logo · caption bottom-left spec (the prompt): comp_id StatScoreboardvisual receiptsshape torusground blacktreatment wireframemotion receipts-countpower receiptsinstrument spotlight→Ask your data at office.temerarii.xyzcurate pillar, stats, statsLabels | |||||||
| 8 |  first render · fix pending shared field signature-3dgeneric_scene | resolve | 158.2–163.2s | coalescence | icon·ember-fill | ♪ bed_out | Social at scale: one story, nine native channels on-screen: Chapter 7 |
expected on screen: black ground · torus hero in the shared Signal Field · Nuntius leads · coalescence · coalescence · icon·ember-fill logo · caption bottom-left spec (the prompt): comp_id shared Signal Fieldvisual coalescenceshape torusground blacktreatment ember-fillmotion coalescencepower coalescenceinstrument coalescence→Chapter 7 | |||||||
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 the data your tools already hold into answers a model can read and defend
This week is about choosing an automation platform with no copy-paste. This points it at business intelligence: turning the data your tools already hold into answers you can act on. Most owners run on a gut feeling because the numbers live in five places. This shows how to pull them into one view a model can read and explain.
What you learn:
- Wire an MCP server into a data source you own, your analytics or your sales sheet, and let the model pull the numbers directly. No exporting, no copy-paste into a slide. The agent queries the live data and writes the summary from what it found.
- Define each number once, in writing: what counts as a lead, what counts as a sale, what date range. Put the definition next to the report. Half of all bad dashboards are two people meaning different things by the same word.
- When the agent gives you a number, make it show the query and the rows it used. A number with no source is a rumor. The point is a number you would defend out loud, not a pretty chart.
- One hard rule: if the data is not there, the answer is I do not know, not a guessed figure. A made-up metric is worse than a blank, because someone will act on it.
We run our own calendar from a single source file for exactly this reason, so the answer is never spread across ten tools we have to reconcile by hand.
Pick the one number you check most, define it in a sentence, and wire a tool straight into the data behind it. Make the model show its work.
See how we run on one source at office.temerarii.xyz.
Keywords: business intelligence, MCP, live data query, metric definitions, no hallucinated numbers, single source of truth