Assemble the FREE steps of the product-hypermotion + kinetic-typography video format — dice ONE Seedance 2.0 hypermotion clip into 5-6 segments and intercut them with PIL kinetic-typography spec/CTA cards (italic skew, 1.08x outline echo, 3D extrusion, slam-with-shake, inversion flash) plus a real-logo PIL end card (base64-decoded from the brand SVG), center-crop 1:1 to 9:16, and explicit-map mux a music bed. Deterministic, FREE (PIL + FFmpeg), no paid calls, the real logo and spec typography stay pixel-crisp. The paid steps (the ONE Seedance i2v, the music) are separate capabilities; the recipe orchestrates them. Use for the product-hypermotion format.
Install gooseworks: https://gooseworks.ai/creative-agent/install.md # then, in Claude Code, Cursor, or Codex: /gooseworks use the render-hypermotion skill
Assembles a product-hypermotion + kinetic-typography ad: a vertical 9:16 sizzle (~20–30s) where ONE spectacular AI-gen hypermotion clip of the hero product carries the energy and punchy PIL-rendered spec cards carry the facts. Music-led, no VO — it reads as a high-production sizzle, not UGC.
The reusable IP is one-call-many-cuts + intercut: dice ONE 12–15s Seedance 2.0 hypermotion i2v into 5–6 segments (never a paid call per segment) and interleave PIL kinetic-typography spec cards between the cuts, capped by a real-logo end card, over a beat-driven music bed.
This capability ships the FREE, deterministic assembly — everything between the two paid
model calls. It is documentation-grade + config: the format has no runnable driver.
scripts/ carries the worked config + the step-by-step pipeline doc; the demo's
reference PIL impls (gen_kinetic_v6.py, gen_endcard_v10.py, assemble_v10.py) are
not vendored; rebuild them per run from PIPELINE.md.
config.json; never its creative values. Card
labels equal their text_cards keys (intro, spec_1..spec_N, cta) so
beat_structure.concat_order resolves to kinetic-movs/<key>.mp4.The calling recipe asks the user these before any paid step; this capability only assembles what they produced. The Glowmule values are a worked example, never defaults.
config.brand_voice, Seedance block 1, music.bpm (demo: adventure, 124 BPM).config.music.prompt (demo: driving outdoor-adventure anthem).The assembly itself is choice-agnostic: segment cut points, card order and the end card read only from config.
-map 0:v -map 1:a
(the default mapping silently ships ~1 kbps garbage audio; verify ffprobe ≈ 192 kbps).# Phase 2 — FREE PIL cards (per config.text_cards + config.end_card)
# render each spec/CTA card + the real-logo end card, 1080x1920, frame-by-frame → mov
# Phase 3 — FREE dice + intercut + mux
# center-crop 1:1→9:16, dice into 5-6 segments, concat in beat_structure order,
# then explicit-map mux the music bed → master-final.mp4Output: master-final.mp4, 1080×1920, ≈20–30s (25s default) h264 (+ 192 kbps aac music) —
14 segments: intro + 6 hypermotion cuts + 5 spec cards + CTA + real-logo end card. All
FREE.
config.brand_voice, Seedance block 1, music.bpm (demo: adventure, 124 BPM).Research supported advertising promises from a specific buyer's context, current product facts, customer language and observed ads. Turn them into static, video or copy ideas; match video ideas to actual recipes and save a product-scoped angle bank for write-video-ad-script. Product-led discovery can use verified facts without reviews. Reference persistence and reach are observations, not proof of performance.
Analyze ad campaign performance data (Google, Meta, LinkedIn) to identify what's working, what's wasting budget, and specific cut/scale/test recommendations. Runs statistical analysis, funnel diagnostics, and multi-channel budget reallocation with specific dollar-amount shift recommendations and scenario modeling.
For paid lead-gen and participant-recruitment ads, replaces vanity CPA with true CAC per qualified lead by joining ad-platform data with downstream funnel events, surfaces tracking gaps, and classifies every creative into Scale / Keep / Investigate / Cut.