Render an 'Instagram-Live social-proof gallery' video from a config — ~5 real brand product stills each framed as an Instagram-LIVE card (IG gradient-ring avatar, username, verified check, red LIVE badge, viewer count, close X, a short claim-free live-comment feed, an empty 'Add a comment…' bar, and reaction hearts floating up the right edge), with a brand-approved benefit sentence building one phrase per slide and a clean logo endcard, rendered deterministically with PIL frames plus FFmpeg — FREE (the music bed comes from create-music-elevenlabs), products and wordmark stay crisp. Use for the ig-live-gallery format.
Install gooseworks: https://gooseworks.ai/creative-agent/install.md # then, in Claude Code, Cursor, or Codex: /gooseworks use the render-ig-live-gallery skill
Render the ig-live-gallery format from a config. Each of the brand's REAL hero product stills is presented as the "live video" inside an authentic Instagram-Live card — dark gradient scrims top & bottom over the bright product, an IG gradient-ring avatar + username + verified check + red LIVE badge + viewer count + close X across the top, a short live-comment feed + an EMPTY "Add a comment…" bar + heart/share icons across the bottom, and pink/red reaction hearts floating up the right edge. A benefit sentence builds one short phrase per slide as the top overlay, and it closes on a clean brand logo card with the payoff line + CTA.
Deterministic and FREE — no generative image/video, no AI-rendered text. The paid music bed is
a separate capability (create-music-elevenlabs) muxed on afterward.
python3 scripts/build.py --config config.json --assets <stills-dir> --out master-silent.mp4
# quick check — one frame per slide:
python3 scripts/build.py --config config.json --assets <stills-dir> --stills <dir>config.example.json is a full worked config (the fictional stationery brand Inkwren) — its phrases, comments, payoff and
music prompt are that demo's creative answers, not defaults. Config shape:
username, verified, palette, slide_dur, endcard_dur, crossfadeslides[] — { image, phrase, viewers, comments: [[handle, text], …] } per productendcard — { logo, logo_is_white, payoff, handle, cta }music — { prompt, length_ms, trim_intro_sec } (consumed by create-music-elevenlabs)Image paths in the config resolve against --assets.
The recipe asks the user these before any paid step; the renderer just draws what the config says.
slide_story — what the slides show and how the sentence builds → slides[].image, slides[].phrase, endcard.payoff. Asked of the user; the demo used a journaling routine, one product per step.copy_tone — voice of the phrases and live comments → slides[].phrase, slides[].comments. Asked of the user; the demo used a cozy, unhurried voice.music — the bed's genre/mood → music.prompt. Asked of the user; the demo used cozy lo-fi keys + bells.L<246 threshold (catches white-product edge
shading too) and centers that midpoint.A silent WxH (default 1080x1920) H.264 master. Generate + mux the create-music-elevenlabs
bed with a short fade in/out + loudnorm (-map 0:v:0 -map 1:a:0), then QC with watch.
username, verified, palette, slide_dur, endcard_dur, crossfadeslides[] — { image, phrase, viewers, comments: [[handle, text], …] } per productendcard — { logo, logo_is_white, payoff, handle, cta }music — { prompt, length_ms, trim_intro_sec } (consumed by create-music-elevenlabs)slide_story — what the slides show and how the sentence builds → slides[].image, slides[].phrase, endcard.payoff. Asked of the user; the demo used a journaling routine, one product per step.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.
A pre-launch check catches cut-off text, low-resolution images and copy that could be rejected. Here is what to look at, and what Goose checks for you before a Meta launch.