Install in seconds
Install this skill
Copy the command and run it in your terminal. You can review the source before installing.
terminal
git clone https://github.com/gooseworks-ai/goose-skills

Works with Git. The repository opens in your current directory.

🎬
AutomationPython

Render Absurdist Explainer

by gooseworks-ai

Deterministic Python+ffmpeg+PIL assembler for absurdist animated-explainer video ads (~38s, 9:16). Retimes clips to VO windows, re-encodes to 30fps, concats, mixes audio with loudnorm, and burns captions using real product imagery for the end card.

1.1K stars194 forksAdded 2026/07/20
claudeclaude-skillsclaudecodeclaudecode-skillscodexcursorgtmskills

Documentation

README

render-absurdist-explainer

The free, deterministic renderer for the absurdist-explainer video ad format β€” the bright Pixar/Disney 3D spot where a personified villain (the problem) narrates the whole ad in one voice, teaches the product's ownable mechanism through cartoon biology, lists the damage, then watches its own scheme collapse when the product arrives. This capability is the FREE assembly stage only. All generative work (nano-banana keyframes, Seedance i2v clips, ElevenLabs VO + music) happens upstream in the recipe and is handed to this capability as files.

It ports the validated compose recipe from two reference runs (HUM "Big Chill" cortisol absurdism and Soteri "Eczema, the pH villain"). The recipe is deterministic β€” iterate the cut for free, re-roll only the offending paid beat.

What it does (the deterministic recipe)

  1. Per-scene retime. Each i2v clip is retimed to its measured VO window (scale=1080:1920:force_original_aspect_ratio=increase,crop=1080:1920,fps=30,setsar=1, then tpad=stop_mode=clone if the VO is longer than the clip, else -t trim).
  2. Identical re-encode. Every segment is re-encoded libx264 -crf 18 -pix_fmt yuv420p -r 30 even if already correct β€” a framerate mismatch makes the concat demuxer silently drop frames.
  3. Concat all scene segments + the end card via the concat demuxer (-c copy).
  4. Real-product end card. build_endcard.py composites the REAL retail product photo over the brand palette (flat, or sampled from the photo's own edge pixel) with a typeset wordmark + claim rows + CTA pill in PIL ImageDraw.text β€” never an AI cartoon bottle, never AI-rendered brand text. compose.py Ken-Burnses it 1.00 β†’ 1.04 over the dwell.
  5. Mix. VO bus loudnorm I=-14 TP=-1.5, music bus loudnorm I=-26 TP=-3 then volume=0.62, amix inputs=2 duration=first normalize=0 β†’ master lands at -14.5..-13.5 LUFS with the music ducked under the VO.
  6. Captions last. make_captions.py emits a libass .ass (one cue per scene, Arial 64 white / 6px outline / MarginV=330, start = scene_start + 0.08s, suppressed on the end card). compose.py burns it as the final filter so captions sit on top.

Scripts (free β€” Python + ffmpeg + PIL, no bash, no paid calls)

  • scripts/build_endcard.py β€” PIL composite of the real product photo + typeset brand layer (wordmark / product line / claim rows / accent CTA pill). Reads the same config.json. Run this FIRST so end_card.image exists before compose.py.
  • scripts/make_captions.py β€” emits the per-scene libass .ass from the SAME scene table compose reads, so caption windows stay in lockstep with the cut. Run before compose.py (or point config.captions_ass at nothing to skip captions).
  • scripts/compose.py β€” the assembler: per-scene retime + identical 30fps re-encode β†’ concat β†’ Ken-Burns end card β†’ VO/music loudnorm mix β†’ burn captions β†’ master mp4.
  • scripts/config.example.json β€” the shape of the config the recipe binds (the brand-neutralised Soteri values as a worked reference).

Inputs (all via --config + a runtime work dir β€” NO hardcoded paths)

config.json carries: scenes[] (each {id, clip, target_sec, vo, caption, atempo?} where target_sec is the measured VO window), end_card{product_image, image, dwell_sec, zoom_to, wordmark, product_line, claims[], cta, background?}, brand_palette {primary, primary_lite, accent, grey}, music_bed, music_volume (default 0.62), atempo (compose-stage VO speed-up, default off; the reference runs used 1.3 when the VO read slow), captions_ass, and caption_style. See config.example.json.

Craft rules (load-bearing β€” faithful to the source molecule)

  • The end card is the REAL product photo, composited β€” never an AI cartoon bottle. Both reference runs shipped an AI bottle first and had to re-shoot with the real photo.
  • No AI-rendered brand text anywhere. Wordmark, claims, CTA, motif β€” all PIL ImageDraw.text. AI draws the world + characters only.
  • Re-encode every segment to 30fps before concat, even if already correct, or the concat demuxer silently drops frames.
  • target_sec is the MEASURED VO duration (ffprobe each VO mp3), never a planned word count β€” VO drives the per-scene timing.
  • Mix constants are validated β€” VO -14 LUFS, music -26 LUFS then volumeβ‰ˆ0.62 (Soteri) to 0.70 (Big Chill), amix normalize=0. Master target -14.5..-13.5 LUFS, true-peak ≀ -1.5 dBFS.
  • Caption start = scene_start + 0.08s, suppressed on the end card (its typeset copy carries the message β€” two text layers at one spot are both unreadable).

Requires

watch (QC the final master β€” confirm the villain silhouette holds, the single voice carries the whole spot, the motif lands β‰₯3Γ—, no AI brand text leaked into a cartoon background, the end card is the real product, and duration is within Β±0.1s of the summed windows). The recipe gates the paid create-image-fal (keyframes), create-video-fal (Seedance i2v), create-vo-elevenlabs, and create-music-elevenlabs calls to their own capabilities β€” this capability itself makes NO paid calls.

More from gooseworks-ai

Other Claude Code skills by this author in the directory.