Motion graphics

Skill rakibulism/agent-skills-os/skills/motion-graphics

THE UNIVERSAL AGENT SKILLS LIBRARY

Install

npx -y skills add rakibulism/agent-skills-os --skill motion-graphics

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Design titles, lower thirds, captions, and animated graphics for video — what moves, when, and why, with legibility and timing that respect the edit. Produces a graphics spec with hierarchy, animation timing/easing, safe-area and legibility notes. Use when the user asks to design titles, lower thirds, captions/subtitles, an intro/outro, kinetic typography, or animated overlays.

SKILL.md

5.6 KB, as published. Nobody here has run it

Motion Graphics

Graphics in video have one job: deliver information without stealing attention from the story. A title that's hard to read, on too long, or animated for its own sake actively hurts the video. Legibility and timing beat flash every time.

Legibility is non-negotiable

If the viewer can't read it in the time it's on screen, it failed — no matter how nice it looks.

Contrast — text must separate from the footage behind every frame it's over. Footage moves; a title that's readable on frame 1 can vanish on frame 30. Use a scrim, drop shadow, subtle stroke, or a solid plate.
Size — readable on a phone, not just your editing monitor. When in doubt, bigger.
Hold long enough to read twice — the rough rule: time to read it aloud, plus a beat. Short titles ~1.5–2s on screen; longer ~3s+.
Limit words — a title card is not a paragraph. Cut to the essential phrase.

Respect the safe areas

Keep text inside the title-safe area (~10% margin from edges) so nothing clips on different displays or gets covered by platform UI (captions, like buttons, progress bars).
For social/vertical (9:16), keep key text out of the top and bottom ~15% where the app's UI sits.
Design for the aspect ratio it ships in, and check reframes (16:9 → 9:16 → 1:1) — text that's centered in one is often cropped in another.

Hierarchy

Every graphic has a primary message and supporting detail. Make the order of reading obvious:

Size, weight, and color establish what's read first. One clear focal point per graphic.
Lower thirds: name = primary (largest/boldest), title/role = secondary (smaller, lighter). Don't make them equal.
Align to a grid. Consistent margins and baselines across all graphics make the whole video feel designed, not assembled.

Animation: motivated, fast, out of the way

Every animation needs a purpose — reveal information, draw the eye, express brand. Motion without a reason is noise. (See interaction-designer for the same principle in UI.)
In and out, both matter. Graphics that snap in and cut out feel cheap; design the exit as deliberately as the entrance.
Stagger related elements (name in, then title in 4–6 frames later) instead of moving everything at once — it reads as intentional and guides the eye.
Ease, don't linear. Ease-out for entrances (fast then settle), ease-in for exits. Linear motion feels robotic. A little overshoot adds life; too much feels gimmicky.
Keep it quick — title animations live around 0.3–0.8s. The animation should be over well before the read time is up; people came to read it, not watch it arrive.
Match the video's energy — punchy cuts → snappy graphics; calm doc → gentle fades.

Common elements

Lower third — animate in from the edge or with a wipe/fade, hold for the read, animate out. Don't leave it up the whole interview.
Title / intro card — strong type, often with the video's hook. Keep brand intros short (<3–4s); long logo stings get skipped.
Kinetic typography — words timed to a voiceover or beat. Powerful but labor-intensive; emphasize key words, don't animate every one.
Callouts / annotations — arrows, highlights, boxes that point at on-screen things. Animate them in when referenced, out when done.
End screen / outro — CTA, subscribe, next video. Leave room and time for the viewer to act; on YouTube, respect the end-screen template zones.

Captions and subtitles

Burned-in captions boost retention and serve sound-off viewers (most social viewing). Style them legibly: high contrast, a plate or stroke, comfortable line length (~1–2 lines, ~32–42 chars/line).
Time to speech — captions appear with the words, not before or lagging behind.
Accessibility — for true subtitles/SDH, include speaker IDs and key non-speech sound cues.

Output format

## Graphics spec: <element>

Format: <aspect/platform>   Appears at: <timeline moment>   On screen: <duration>

### Layout
- Primary: <text> — <font/size/weight/color>
- Secondary: <text> — <…>
- Background: <scrim/plate/shadow for legibility over moving footage>
- Safe area: <margins, platform-UI clearance>

### Animation
| Element | In (timing/easing) | Hold | Out |
|---|---|---|---|
| <name> | fade+rise 0.4s ease-out | 2.0s | fade 0.3s ease-in |

### Cautions
- <contrast over busy footage, reframe crops, read-time, brand fonts>

Always check the graphic against the actual footage behind it across its whole duration and against the shipping aspect ratio — most motion-graphics failures are legibility-over-moving-background or text cropped on a reframe.