render-3d-character-explainer

Name: render-3d-character-explainer
Author: gooseworks-ai

1.0kstars193forks

How to install render-3d-character-explainer?

gooseworks-ai/goose-skills/render-3d-character-explainer

$npx -y skills add gooseworks-ai/goose-skills --skill render-3d-character-explainer

Installs into the current project.

Summary

Assemble a glossy 3D-character animated-explainer video ad (~77s, 9:16) built on an "N types of X" listicle spine — a recurring human protagonist plus a locked cast of N persona characters, one per list item. Given the per-scene i2v clips + a per-scene target-duration table + a n

Files · 1

View on GitHub

SKILL.md

1	# render-3d-character-explainer
2
3	The free, deterministic renderer for the 3d-character-explainer video ad format — the
4	glossy Pixar-style 3D spot built on an "N types of X" listicle spine, where a recurring
5	human protagonist plus a locked cast of N persona characters (one per list item) carry a
6	hook → "deeper story" → cast-reveal → one beat per list item → kicker → product test →
7	relieved payoff. This capability is the FREE assembly stage only. All generative work
8	(Nano-Banana cast anchors + per-scene keyframes, Kling-V3 i2v clips, ElevenLabs VO + music,
9	or a source ad's audio reused verbatim) happens upstream in the recipe and is handed to this
10	capability as files.
11
12	It ports the validated compose recipe from the Bristle "Six Types" restyle run
13	(`_render_full.sh` — per-scene trim → normalize 1080×1920/fps30 → concat -c copy → mux the
14	source audio, with a static-still fallback on any failed clip). The assembly is
15	deterministic — iterate the cut for free, re-roll only the offending paid beat.
16
17	## Two modes
18
19	- Restyle mode (`audio_mode: "restyle"`, the reference run) — re-tell a finished source
20	ad, beat for beat, as 3D character comedy. The source ad's **audio mix (VO + music bed) is
21	reused VERBATIM** (`source_audio`), and the per-scene `target_sec` table is inherited from
22	the source's scene timing. No new VO or music is rendered. The trims must sum to the source
23	audio length.
24	- Original mode (`audio_mode: "original"`) — the ad authors its own narration. Each scene
25	carries a measured VO cue (`scenes[].vo`, `target_sec` = the ffprobe'd VO duration) which is
26	concatenated into a VO track (loudnorm I=-14) and optionally mixed under a `music_bed`
27	(loudnorm I=-26 then `volume`, `amix normalize=0`).
28
29	## What it does (the deterministic recipe)
30
31	1. Per-scene retime. Each i2v clip is trimmed to its scene `target_sec` and normalized to
32	identical dims/fps/SAR
33	(`scale=W:H:force_original_aspect_ratio=decrease,pad=W:H:(ow-iw)/2:(oh-ih)/2:color=<pad>,fps=30,setsar=1`).
34	A clip shorter than its window is extended with `tpad=stop_mode=clone`; a longer one is
35	`-t` trimmed. Decrease+pad (never crop) preserves the full 9:16 keyframe framing.
36	2. Static-still fallback. For any scene whose `clip` is missing or failed to render, the
37	scene's `keyframe` PNG is looped (`-loop 1`) for `target_sec`, so the master always
38	assembles. Fallback scenes are printed at the end.
39	3. Identical re-encode + concat. Every segment is re-encoded `libx264 -crf 18 -pix_fmt
40	yuv420p -r 30` even if already correct — a dims/framerate mismatch makes the concat demuxer
41	silently drop frames — then concatenated via the concat demuxer (`-c copy`).
42	4. Audio. Restyle: `source_audio` muxed verbatim (`-map 0:v -map 1:a`), clamped to the
43	video length. Original: per-scene VO track (optional `atempo`, `apad`, `-t` clamp) → loudnorm
44	→ optionally mixed under the music bed.
45	5. Captions last. `make_captions.py` emits a libass `.ass` (one cue per scene, `start =
46	scene_start + 0.08s`, suppressed on any scene with no caption — e.g. a product/end-card beat
47	carrying its own typeset copy). `compose.py` burns it as the final filter so captions sit on
48	top. Word-level energy-pop captions (Whisper on the narration) are the recipe's upstream
49	option — produce that `.ass` externally and point `captions_ass` at it; compose burns
50	whatever `.ass` it's handed.
51
52	## Scripts (free — Python + ffmpeg, no bash, no paid calls)
53
54	- `scripts/make_captions.py` — emits the per-scene libass `.ass` from the SAME scene table
55	compose reads, so caption windows stay in lockstep with the cut. Run before `compose.py`
56	(or leave `captions_ass` unset / pointing at nothing to skip captions).
57	- `scripts/compose.py` — the assembler: per-scene t

People who liked this also viewed

6 picks

Type

render-3d-character-explainer

1.0kstars193forks

View on GitHub

How to install render-3d-character-explainer?

gooseworks-ai/goose-skills/render-3d-character-explainer

$npx -y skills add gooseworks-ai/goose-skills --skill render-3d-character-explainer

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# render-3d-character-explainer
2
3	The free, deterministic renderer for the 3d-character-explainer video ad format — the
4	glossy Pixar-style 3D spot built on an "N types of X" listicle spine, where a recurring
5	human protagonist plus a locked cast of N persona characters (one per list item) carry a
6	hook → "deeper story" → cast-reveal → one beat per list item → kicker → product test →
7	relieved payoff. This capability is the FREE assembly stage only. All generative work
8	(Nano-Banana cast anchors + per-scene keyframes, Kling-V3 i2v clips, ElevenLabs VO + music,
9	or a source ad's audio reused verbatim) happens upstream in the recipe and is handed to this
10	capability as files.
11
12	It ports the validated compose recipe from the Bristle "Six Types" restyle run
13	(`_render_full.sh` — per-scene trim → normalize 1080×1920/fps30 → concat -c copy → mux the
14	source audio, with a static-still fallback on any failed clip). The assembly is
15	deterministic — iterate the cut for free, re-roll only the offending paid beat.
16
17	## Two modes
18
19	- Restyle mode (`audio_mode: "restyle"`, the reference run) — re-tell a finished source
20	ad, beat for beat, as 3D character comedy. The source ad's **audio mix (VO + music bed) is
21	reused VERBATIM** (`source_audio`), and the per-scene `target_sec` table is inherited from
22	the source's scene timing. No new VO or music is rendered. The trims must sum to the source
23	audio length.
24	- Original mode (`audio_mode: "original"`) — the ad authors its own narration. Each scene
25	carries a measured VO cue (`scenes[].vo`, `target_sec` = the ffprobe'd VO duration) which is
26	concatenated into a VO track (loudnorm I=-14) and optionally mixed under a `music_bed`
27	(loudnorm I=-26 then `volume`, `amix normalize=0`).
28
29	## What it does (the deterministic recipe)
30
31	1. Per-scene retime. Each i2v clip is trimmed to its scene `target_sec` and normalized to
32	identical dims/fps/SAR
33	(`scale=W:H:force_original_aspect_ratio=decrease,pad=W:H:(ow-iw)/2:(oh-ih)/2:color=<pad>,fps=30,setsar=1`).
34	A clip shorter than its window is extended with `tpad=stop_mode=clone`; a longer one is
35	`-t` trimmed. Decrease+pad (never crop) preserves the full 9:16 keyframe framing.
36	2. Static-still fallback. For any scene whose `clip` is missing or failed to render, the
37	scene's `keyframe` PNG is looped (`-loop 1`) for `target_sec`, so the master always
38	assembles. Fallback scenes are printed at the end.
39	3. Identical re-encode + concat. Every segment is re-encoded `libx264 -crf 18 -pix_fmt
40	yuv420p -r 30` even if already correct — a dims/framerate mismatch makes the concat demuxer
41	silently drop frames — then concatenated via the concat demuxer (`-c copy`).
42	4. Audio. Restyle: `source_audio` muxed verbatim (`-map 0:v -map 1:a`), clamped to the
43	video length. Original: per-scene VO track (optional `atempo`, `apad`, `-t` clamp) → loudnorm
44	→ optionally mixed under the music bed.
45	5. Captions last. `make_captions.py` emits a libass `.ass` (one cue per scene, `start =
46	scene_start + 0.08s`, suppressed on any scene with no caption — e.g. a product/end-card beat
47	carrying its own typeset copy). `compose.py` burns it as the final filter so captions sit on
48	top. Word-level energy-pop captions (Whisper on the narration) are the recipe's upstream
49	option — produce that `.ass` externally and point `captions_ass` at it; compose burns
50	whatever `.ass` it's handed.
51
52	## Scripts (free — Python + ffmpeg, no bash, no paid calls)
53
54	- `scripts/make_captions.py` — emits the per-scene libass `.ass` from the SAME scene table
55	compose reads, so caption windows stay in lockstep with the cut. Run before `compose.py`
56	(or leave `captions_ass` unset / pointing at nothing to skip captions).
57	- `scripts/compose.py` — the assembler: per-scene t

People who liked this also viewed

6 picks

Type