Iterating

Editing existing videos already leaned on two habits — one variable per render, absolute targets over relative nudges — for every verb on that page. This is where they come from, stated plainly, so you can apply them anywhere: HyperFrames is a conversation. After the first render, talk to the agent the way you’d talk to a video editor — don’t re-prompt from scratch:

Make the title 2x bigger.

Swap to dark mode.

Add a fade-out at the end and a lower third at 0:03 with my name and title.

The captions are too small and they overlap the lower third. Move them up and shrink them.

Replace the background music with assets/track.mp3.

The agent already has the composition open and the skills loaded — small targeted edits produce better results than long re-specifications.

​

The gates can’t tell you it’s good

Before the loop: know what you are iterating against. lint and check are necessary and not sufficient — they verify structure, not quality. lint reads the HTML, check runs the composition in a headless browser and reports runtime errors, layout collisions, motion problems, and contrast failures. Every one of those is a fact about whether the composition is well-formed. None of them is an opinion about whether the video is good. So the failure mode to expect is: passes every automated gate, fails human review. Both earlier builds of this guide’s capstone did exactly that — one came back clean and read as sparse elements floating in empty frames, the next came back clean and read as a well-dressed slideshow. No linter has an opinion about either. Which is why the loop below always ends at the render, not at the gate. Watch the output, name what’s wrong in specific terms, and change one thing.

​

Editing is for nudging, not gambling

Three moves keep iteration convergent instead of circular:

1. One variable per edit. Change one thing, render, judge, repeat. A request that moves three things at once (“bigger title, warmer colors, faster cuts”) makes it impossible to attribute what helped and what hurt.
2. Layer one element type at a time. Building up a complex scene? Start with the minimal version (subject + its motion only), confirm it reads, then add exactly one layer per pass — camera move, then background motion, then style treatment. Every addition stays attributable.
3. Strip, then re-layer. When a scene keeps misfiring, don’t pile on corrections — strip it back to the simplest version that works (freeze the camera, simplify the motion, clear the background), confirm, then reintroduce complexity one layer at a time until you find the ingredient that breaks it.

​

Calibrate with absolute targets

When you’re dialing in a look, relative corrections oscillate — “make the dots finer” overshoots, “a bit bigger” overshoots back. State targets as absolute values and the agent lands them in one pass:

• ❌ make the dots 2x finer
• ✅ dot radius = 25% of row spacing, with clear gaps between dots
• ❌ the glow is too strong
• ✅ glow at ~35% peak opacity, always subtler than the orb itself

And freeze what already works: “the framing and motion are right — don’t touch them; only change the dot size.” Without the freeze clause, a rebuild can drift on axes you’d already settled. This is the clause in the full capstone prompt that buys the piece — prompt language you can lift for your own video:

The chip sits ON the wire at every moment: bind its position to the wire’s actual path geometry (sample the drawn path, seek-safe), never keyframe it independently — through every bend, arc, waveform, and dive, the chip visibly hugs the line. A chip floating off-path breaks the story.

Next: Recreating something you saw — the same absolute-target discipline, applied to matching a specific reference frame by frame.