gooseworks-ads-skills

Produce a Photo-Grid promo card — a silent, music-optional feed-placement video where a scrolling 2×N tile grid (product photos + short brand video clips + big-type offer tiles + a promo-code tile) animates under a centered wordmark, headline, sub, and outline feature chips. 9:16 or 4:5, ~10s, $0 HTML-hyperframe pipeline (no T2V, no VO). Recreates the Meta "photo-grid mosaic" template (e.g. "Our New Band 2.0 / LAUNCH30"). Validated on Hume Health (Father's Day promo, 2026-06-03). Use when the brand wants a clean editorial feed card that works muted; NOT for kinetic-BG vignette (use create-vignette-ad), UGC, or VO-driven ads.

Remix a published Apple Notes video ad from the Goose Ads library for a new brand — keep the source's note structure and typing pacing, swap in the brand's product, voice, and promo code, get the note approved in-session, then render with create-apple-notes-video-ad and publish the MP4 back to Gooseworks with live stage reporting. The Apple Notes typing-reveal counterpart to remix-imessage-ad-from-sample; this is what the app's Apple Notes format tab calls.

Remix a published ChatGPT video ad from the Goose Ads library for a new brand — keep the source's conversation structure and pacing (the single ask → streamed answer beat), swap in the brand's product as the natural recommendation and the promo code woven into the response, get the conversation approved in-session, then render with create-chatgpt-video-ad and publish the MP4 back to Gooseworks with live stage reporting. The ChatGPT chat-reveal / ask-ChatGPT counterpart to remix-imessage-ad-from-sample; this is what the app's ChatGPT format tab calls.

Remix a published iMessage video ad from the Goose Ads library for a new brand — keep the source's conversation structure and pacing, swap in the brand's product, voice, and promo code, get the thread approved in-session, then render with create-imessage-video-ad and publish the MP4 back to Gooseworks with live stage reporting. The video counterpart to remix-graphic-ad-from-reference; this is what the app's iMessage format tab calls.

Remix a published Photo-Grid promo card from the Goose Ads library for a new brand — keep the source's grid layout, tile rhythm, scroll, and offer structure, swap in the brand's product photos, corpus video clips, wordmark, palette, copy, and promo code, get the card approved in-session, then render with create-photo-grid-ad and publish the MP4 back to Gooseworks with live stage reporting. The photo-grid counterpart to remix-imessage-ad-from-sample; this is what the app's Photo-Grid format tab calls.

mix a music track under a silent video using ffmpeg. normalizes loudness to roughly -14 LUFS, applies a short fade in and out, and trims or loops the music to match video duration. use after a silent video master and a chosen music file are both ready.

Burn 2–3-word, beat-synced lyric captions into a music-video ad. Reads word-level timestamps (ElevenLabs music metadata or Whisper transcript of the rendered song), chunks lyrics into 1–3 word bursts, emphasizes operator-tagged accent words, and produces an ASS overlay plus a timeline preview strip.

create product, avatar, b-roll, or stop-motion stills using nano banana style prompts.

Produce a 9:16 social-native ad recreating the iPhone Apple Notes typing experience — the note begins with 1–2 visible lines, then progressively types additional paragraphs character-by-character with per-keystroke iPhone keyboard SFX. One continuous Playwright recording, music bed underneath, brand end card with CTA. Assembled to a master MP4 plus a 9:16 social export.

Produce a 9:16 social-native ad that recreates a ChatGPT mobile chat — user types in the composer with the iOS keyboard visible, taps send, keyboard slides down, header right-cluster swaps (personPlus/dotted → edit/more), one gray loading dot appears, then the assistant response streams in word-by-word with a soft opacity ramp. Continuous Playwright recording, subtle SFX (key-tap / send-tap / stream-tick / response-done), assembled into a master MP4 plus 9:16 / 1×1 exports. No music bed and no end card by default.

Remix a published cinematic music-video ad from the Goose Ads library for a new brand — keep the source's 14-tableau 3-act arc, cut-on-vocal-beat structure, and one-look-pack discipline, swap in the brand's product as the object-hero, a look pack that fits the brand, original anthem lyrics in the brand's campaign voice, and the brand's wordmark + tagline on the end card, get the storyboard approved in-session, then render with create-cinematic-music-video and publish the MP4 back to Gooseworks with live stage reporting. The generative, paid sibling of remix-imessage-ad-from-sample; this is what the app's music-video format tab calls.

Research a brand and produce a clean, reusable brand-context pack that any downstream video/ad skill can consume with zero re-work — four structured markdown docs (brand-summary, visual-identity, competitors, audience), a dated source list, an optional UI-reference doc, sourced logo + reference photos, an optional set of brand-anchored generated stills, and a concept brief. Every binary asset is cataloged in brand-assets/manifest.json with a name + usage description so an agent picks assets by purpose, not filename. Use when starting work on a brand a project hasn't touched before, or when a brand folder has empty/stub research that needs filling. Core path needs no API keys (runs on the agent's web tools); only the optional image step needs a FAL key. Walks an 8-phase pipeline (disambiguate → scaffold → web research → existing-ad analysis → source assets → optional imagery → concept brief → verify) with human gates, and ends on a deterministic verify_pack.py check.

Render pixel-accurate Apple Notes (iPhone, light mode) screenshot mockups from a JSON note spec. Outputs HTML + PNG at the iPhone 16/15 Pro native 1180×2556. Supports paragraphs, images, checklists, dividers, autocorrect underline, smart quotes, and an optional iOS keyboard chrome overlay used by the parent video-ad molecule.

Render pixel-accurate ChatGPT mobile (iOS) screen mockups in light mode from a thread JSON. Supports user text bubbles, user image attachments, assistant markdown prose, citation chips, the OpenAI spiral logo, the Apps-SDK GPT chip in the composer, and three header styles (model-tag, plain title, "Get Plus"). Fixed 9:16 viewport. Outputs HTML + PNG.

Render pixel-accurate iMessage screenshot mockups (DM or group) from a thread JSON. Supports minimal, with-keyboard, and full iPhone 15 Pro frame variants. Outputs HTML + PNG.

Render a static iOS QWERTY keyboard as inline HTML+CSS sized for the 750-wide 9:16 stage. Includes the 3-suggestion bar, alpha keys, shift/backspace, 123/emoji/space/return row, and the bottom globe+mic strip. No animation logic — slide up/down is the molecule's job (CSS transform on the .keyboard root).

Produce a 9:16 social-native ad that recreates an iMessage conversation reveal — bubbles pop in over time, composer types char-by-char, real Apple iMessage SFX hit on every send/receive, music bed underneath, brand end card with FREEPACK-style CTA. One continuous Playwright recording (no scene-cuts/glitches) in plain-chat or iPhone-framed (Dynamic Island over a background) variants, assembled to a master MP4 plus 9:16 / 1×1 export variants.

Recreate a static graphic ad (Pinterest pin, IG/FB feed image, poster) from a reference image, swapping in a new brand's product and new copy while keeping the reference's layout, composition, and visual energy. Two engines — HTML overlay (default; editable pixel-perfect text, free) and GPT Image 2 in edit-the-reference mode (only when a real photographic element is needed). The static-graphics counterpart to the video remix-ad skill; this is what the app calls when a user picks a reference ad and wants it for their own product.

Write the ad SCRIPT for a remix — take the brand details (a brand-context pack and/or a source ad's remix payload) plus the user's creative prompt, and produce a structured, render-ready script in the **format the source actually is** — podcast-skit, iMessage thread, sung music video, single-host UGC, animated explainer, etc. In remix mode the skill **watches the source video first** (via `/watch`), reads `source-sample.json.formatProfile`, and **routes to the matching script shape and production template — it does NOT default to `podcast-script`**. Only `podcast-skit-fabricated` has a shipped scaffolder today; every other format is hand-authored against the source's actual scene shape. The skill then asks the user format-aware clarifying questions (angle, references, format-specific knobs, must-include / must-avoid), writes the lines (or bubbles / lyrics / VO beats), emits a structured `storyboard.json` (per-scene stills + optional rendered keyframe paths) the consuming app reads to **preview keyframes** wit

Given the path to a finished content-goose ad-run folder, extract everything that defines that ad — recipe shot list, VO script, characters, voices, world, atom-skills, master mp4 — and emit a `source-sample.json` in the exact shape the `upload-ad-sample` skill writes to the Goose Ads library. Also links every character and voice to the central character library at `/Users/akhil/projects/content-goose/assets/character-library/`, and if a character isn't in the library yet, adds it first then links. Use when the user wants to remix one of their existing ads — this skill produces the source JSON that the script-rewriting step and `remix-ad` consume.

Remix an ad already uploaded to the Goose Ads library into a NEW ad — same proven structure, retargeted to a different brand/product (or the same brand, new angle). Reads the sample's remix payload (recipe shot-list, extracted_script, how_to, and remix_spec of atom-skills / voices / worlds / characters), decides what to KEEP vs SWAP, rewrites the script for the new brand while preserving the beat structure and timing, then routes to the matching production template (animated-explainer-ad, create-podcast-skit-ad, …) and runs the kept atom-skills in recipe order. The consumer counterpart to the upload-ad-sample skill. Use when the user says "remix this ad", "make this for [other brand]", "same format, new brand", or references a library sample to riff on. Needs a Goose Ads API token; regeneration needs the routed template's keys.

deterministically execute an edl.json into a finished master.mp4 — the counterpart to generate-edit-decision-list. Resolves file (conform) and generate-mode clips, applies speed/motion, concatenates as hard cuts, mixes audio bed (music + VO + SFX + silence beats + ducking), burns ASS captions, and writes manifest.json. Use as the edit-video step in the orchestrator and every archetype molecule.

stitch video segments with ffmpeg concat, xfade, overlay, audio mux, and export settings.

remove or replace image backgrounds for product assets.

add sound effects or asmr-style audio details to a video mix.

detect beats, bars, downbeats, onsets, and BPM from an audio or video file and write a reusable beat-grid.json. Use when you need to snap edit cuts, captions, or SFX to a musical grid — feeds generate-edit-decision-list, render-edl, and review-video-pacing-rhythm.

mute precise audio windows in a video or audio file to create deliberate silence beats — the tension-beat / act-break device from CREATOR_GRAMMAR.md §6. Does not change video length or frame content; only zeros the audio within the specified windows. Use when an EDL calls for `audio_beats[].type = "silence-beat"` or when you want to punch in an audio dropout before a reveal.

Canonical short-form-ad audio mix in one FFmpeg pass. VO loudnorm + 3.0× per-clip + 2.0× mix, music 0.13 base + apad+afade, sidechain compress 20:1 @ 0.01, climax line +20%, optional video-to-music duration sync. Replaces the reactive multi-round tuning that cost v03 6+ passes.

Two-pass ffmpeg loudnorm. Brings any video or audio file to a target integrated LUFS with true-peak ceiling and LRA cap. Default is -14 LUFS (social).

synthesize sound effects (bell, thud, punch, whoosh, riser, sub-bass drop, breath, droplet) from scratch with ffmpeg sine waves + decay envelopes. Free, deterministic, no API dependency.

burn local ass captions or klap-style captions into a video.

add final klap captions to an existing video using the klap api workflow.

Burns word-level captions into a video in one of several house styles. Transcribes via fal.ai Whisper, builds per-style cues, renders via libass (ffmpeg) or Remotion, and emits .ass/.srt sidecars next to the captioned video. Use when the user asks to caption a video, add subtitles, add karaoke captions, add burned-in text, or repurpose a clip for social.

Snap SRT cue starts to whisper word-level timestamps within ±100 ms so captions land on word boundaries instead of drifting.

design and render caption styles for video outputs. supports six named presets from caption-style-presets.json (bold-tiktok-hook, serif-wellness, handwritten-educator, mono-proof, pinned-comment-card, platform-default) as well as free-form ASS generation and caption PNG rendering. use when a video pipeline needs styled ASS subtitles, title-card PNGs, or captioned overlays.

find licensed stock clips on pexels for use in video edits.

Apply uniform color grade across all clips in an ad run so the polished cut has consistent contrast, saturation, and gamma. Reads a reference clip (default first scene) and matches every other clip's mean LAB stats to it.

apply punch-in, push-in, or Ken Burns zoom motion to a video clip or still image using ffmpeg crop/scale/zoompan. Use when a clip needs a creator-native emphasis move — an instant stepped tighter crop (zoom-punch), a gradual scale-up (push-in), or a slow pan-and-zoom (ken-burns).

compose two video clips side-by-side or stacked into a single 9:16 (or any aspect) frame using ffmpeg. Use when you need the creator-native split-screen layout — e.g. pairing a reaction with a product demo, or stacking a talking-head over b-roll — and want a generic, format-agnostic compositor that is not tied to a specific ad template.

remove pauses and dead air from a talking-head clip by detecting silence gaps and deleting them, producing a tighter re-encoded video. Use when you have a raw UGC or talking-head take and want creator-native breath-removal pacing — either from word-timestamp JSON (transcribe-audio-fal output shape) or via automatic ffmpeg silencedetect.