render-vo-anchored-motion-listicle

Name: render-vo-anchored-motion-listicle
Author: gooseworks-ai

1.0kstars193forks

How to install render-vo-anchored-motion-listicle?

gooseworks-ai/goose-skills/render-vo-anchored-motion-listicle

$npx -y skills add gooseworks-ai/goose-skills --skill render-vo-anchored-motion-listicle

Installs into the current project.

Summary

Assemble an expert/educator motion-graphic LISTICLE video ad from a config — a spoken authoritative voiceover carries a numbered listicle while N web-animated hyperframe beats (HTML plus the Web Animations API, one branded design system of alternating tiles, big hero numerals, an

Files · 1

View on GitHub

SKILL.md

1	# render-vo-anchored-motion-listicle
2
3	Assemble an expert/educator motion-graphic listicle ad from a config: an authoritative spoken
4	voiceover carries a numbered listicle (hook + N points + CTA) and every visual beat is anchored to
5	the VO's word-level timestamps. Each beat is a web-animated hyperframe (an HTML page + the Web
6	Animations API driven by `window.renderAt(t)`) rendered to video frame-by-frame with Playwright, all
7	in ONE branded design system (alternating background tiles, big hero numerals, body type, decorative
8	SVG accents, glass-pill callouts). Periodic color-graded B-roll windows give visual breath, and
9	captions burn only on the B-roll windows. The shipped master is pure motion-graphic + VO — there
10	is NO lipsync (the still expert headshot is kept only for a future lipsync variant). This capability
11	is the FREE, deterministic assembly — the Playwright beat render, the ffmpeg concat, the
12	window-masked caption burn, the VO+music mix, and the final composite.
13
14	`scripts/config.example.json` is the worked example (Everself "doctor-educator" listicle, ~66s
15	1080×1920 9:16 at 25fps); `scripts/PIPELINE.md` maps every config block to its source step and
16	`scripts/README.md` documents the free assembly.
17
18	## Run
19
20	This is the FREE, deterministic assembly stage — it spends nothing. The paid inputs are separate
21	capabilities — the spoken VO (`create-vo-elevenlabs`, a cloned or cast expert voice, `eleven_v3` +
22	`atempo`) whose word-level timestamps (Groq `whisper-large-v3` word-level) set the timeline; the low
23	music bed (`create-music-elevenlabs`); and the stock B-roll (`media-proxy`, trimmed + color-graded).
24	Given the VO + `words-flat.json` + the N authored hyperframe beats + the color-graded B-roll windows +
25	the brand wordmark SVG, `render-vo-anchored-motion-listicle` renders each beat frame-by-frame via
26	Playwright (all beats at fps 25), concats the beats + B-roll, burns the window-masked captions, mixes
27	the VO under the low music bed, and composites → the master. Re-cuts reuse the existing VO / beats /
28	B-roll and cost $0.
29
30	## Contract (the free assembly)
31
32	- The spoken VO carries the listicle — it sets the timeline. The cloned/cast expert VO is the
33	spine; Whisper-transcribe it to word-level timestamps and anchor every beat reveal to those word
34	times. There is no on-camera human and NO lipsync in the shipped master.
35	- Beats are web-animated hyperframes, rendered deterministically. Each beat is an HTML page + the
36	Web Animations API driven by `window.renderAt(t)`; Playwright screenshots it frame-by-frame and
37	ffmpeg encodes it. This is NOT i2v — it is deterministic web motion graphics.
38	- ALL beats share the same fps (25). A mismatched-fps beat stutters at the concat seam. Render
39	every beat and every B-roll window at fps 25.
40	- ONE design system across every beat. A single `_shared.css` (palette + type + alternating
41	tiles + accents + glass-pill) so N beats read as one designed reel; alternate only the background
42	tile, keep numerals / body / accents / pills consistent.
43	- Periodic color-graded B-roll windows for breath. Trim + color-grade each stock/brand clip to
44	the palette (fps 25). These windows are the ONLY captioned windows.
45	- Captions ONLY on the B-roll windows. On the motion-graphic beats the on-screen type IS the
46	caption — burning Whisper captions there double-stacks text. Build the caption ASS from the VO word
47	timings, kept only inside the B-roll windows, 2-word chunks, closing a cue on any >0.4s word gap.
48	The ASS `Format:` header MUST carry a `Name` field — without it the leading-comma bug eats the
49	first field and captions silently drop. If the host ffmpeg lacks libass, render the cues as timed
50	PIL PNG overlays (ffmpeg `overlay=…:enable='between(t,st,en)'`) at the same placement.
51	- Low music bed under the VO. The VO is the spine; the ElevenLabs Music bed sits ~0.18 vol under
52	it. No ducking needed at that level.
53	- **Never AI-

People who liked this also viewed

6 picks

Type

render-vo-anchored-motion-listicle

1.0kstars193forks

View on GitHub

How to install render-vo-anchored-motion-listicle?

gooseworks-ai/goose-skills/render-vo-anchored-motion-listicle

$npx -y skills add gooseworks-ai/goose-skills --skill render-vo-anchored-motion-listicle

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# render-vo-anchored-motion-listicle
2
3	Assemble an expert/educator motion-graphic listicle ad from a config: an authoritative spoken
4	voiceover carries a numbered listicle (hook + N points + CTA) and every visual beat is anchored to
5	the VO's word-level timestamps. Each beat is a web-animated hyperframe (an HTML page + the Web
6	Animations API driven by `window.renderAt(t)`) rendered to video frame-by-frame with Playwright, all
7	in ONE branded design system (alternating background tiles, big hero numerals, body type, decorative
8	SVG accents, glass-pill callouts). Periodic color-graded B-roll windows give visual breath, and
9	captions burn only on the B-roll windows. The shipped master is pure motion-graphic + VO — there
10	is NO lipsync (the still expert headshot is kept only for a future lipsync variant). This capability
11	is the FREE, deterministic assembly — the Playwright beat render, the ffmpeg concat, the
12	window-masked caption burn, the VO+music mix, and the final composite.
13
14	`scripts/config.example.json` is the worked example (Everself "doctor-educator" listicle, ~66s
15	1080×1920 9:16 at 25fps); `scripts/PIPELINE.md` maps every config block to its source step and
16	`scripts/README.md` documents the free assembly.
17
18	## Run
19
20	This is the FREE, deterministic assembly stage — it spends nothing. The paid inputs are separate
21	capabilities — the spoken VO (`create-vo-elevenlabs`, a cloned or cast expert voice, `eleven_v3` +
22	`atempo`) whose word-level timestamps (Groq `whisper-large-v3` word-level) set the timeline; the low
23	music bed (`create-music-elevenlabs`); and the stock B-roll (`media-proxy`, trimmed + color-graded).
24	Given the VO + `words-flat.json` + the N authored hyperframe beats + the color-graded B-roll windows +
25	the brand wordmark SVG, `render-vo-anchored-motion-listicle` renders each beat frame-by-frame via
26	Playwright (all beats at fps 25), concats the beats + B-roll, burns the window-masked captions, mixes
27	the VO under the low music bed, and composites → the master. Re-cuts reuse the existing VO / beats /
28	B-roll and cost $0.
29
30	## Contract (the free assembly)
31
32	- The spoken VO carries the listicle — it sets the timeline. The cloned/cast expert VO is the
33	spine; Whisper-transcribe it to word-level timestamps and anchor every beat reveal to those word
34	times. There is no on-camera human and NO lipsync in the shipped master.
35	- Beats are web-animated hyperframes, rendered deterministically. Each beat is an HTML page + the
36	Web Animations API driven by `window.renderAt(t)`; Playwright screenshots it frame-by-frame and
37	ffmpeg encodes it. This is NOT i2v — it is deterministic web motion graphics.
38	- ALL beats share the same fps (25). A mismatched-fps beat stutters at the concat seam. Render
39	every beat and every B-roll window at fps 25.
40	- ONE design system across every beat. A single `_shared.css` (palette + type + alternating
41	tiles + accents + glass-pill) so N beats read as one designed reel; alternate only the background
42	tile, keep numerals / body / accents / pills consistent.
43	- Periodic color-graded B-roll windows for breath. Trim + color-grade each stock/brand clip to
44	the palette (fps 25). These windows are the ONLY captioned windows.
45	- Captions ONLY on the B-roll windows. On the motion-graphic beats the on-screen type IS the
46	caption — burning Whisper captions there double-stacks text. Build the caption ASS from the VO word
47	timings, kept only inside the B-roll windows, 2-word chunks, closing a cue on any >0.4s word gap.
48	The ASS `Format:` header MUST carry a `Name` field — without it the leading-comma bug eats the
49	first field and captions silently drop. If the host ffmpeg lacks libass, render the cues as timed
50	PIL PNG overlays (ffmpeg `overlay=…:enable='between(t,st,en)'`) at the same placement.
51	- Low music bed under the VO. The VO is the spine; the ElevenLabs Music bed sits ~0.18 vol under
52	it. No ducking needed at that level.
53	- **Never AI-

People who liked this also viewed

6 picks

Type