Communitygithub.com

alecs5am/ralphy

Full end-to-end workflow for generating a realistic AI UGC ad for a product. 15-second format. Looks real. Doesn't feel AI.

¿Qué es ralphy?

ralphy is a Claude Code agent skill that full end-to-end workflow for generating a realistic AI UGC ad for a product. 15-second format. Looks real. Doesn't feel AI.

Compatible con✓Claude Code~Codex CLI~Cursor✓Gemini CLI
npx skills add https://github.com/alecs5am/ralphy/tree/HEAD/notes/skills/ugc-ad-production
Ver original→Ver todas las habilidades

Preguntar en tu IA favorita

Abre un nuevo chat con esta habilidad de agente ya precargada.

OpenAIChatGPTClaudeClaudeGeminiGeminiGrokGrokPerplexityPerplexityDeepSeekDeepSeek

Documentación

UGC Ad Production Pipeline

Required Inputs (ask the user for ALL of these before starting)

Before generating anything, collect:

  • Product — what is being advertised (name, URL, or image)
  • Reference UGC video — a real UGC video in the same niche/style (Pinterest, TikTok, YouTube link). Used for script style, creator personality, and shot reference.
  • Creator face references — 2+ images of real people (attractive, matching the brand vibe). These get face-mixed in Nano Banana Pro to create a new, non-existent creator.
  • Target audience — who the ad is for (e.g. women 18-35 with acne-prone skin)
  • Hook type — problem/solution, before/after, testimonial, transformation, or let the agent decide based on the product
  • NEVER ask
  • Which editing tool to use (always Canva)
  • Video length (always 15 seconds)
  • Aspect ratio (always 9:16 vertical)
  • Which video model (always Kling 3.0)
  • Which image model (always Nano Banana Pro)
  • Step 1 — Script Writing (Claude 4.6)
  • Tool: Claude claude-sonnet-4-6 (strong reasoning + creative writing — feed it as much context as possible)

Prompt template:

Make a UGC script for a 15-second video like [REFERENCE VIDEO] but for [PRODUCT]. Use [CREATOR DESCRIPTION from face references] as the UGC creator/speaker. Include:

  • Hook (first 3-5 seconds): show the problem visually + audio hook
  • Voiceover lines with exact words
  • Cut descriptions: what the camera shows at each moment (creator face, product, before/after, etc.)
  • Actions and mannerisms: make the creator feel like a real character — nervous laugh, hair tuck, direct eye contact, pointing at product, etc.
  • Call to action (last 2-3 seconds)
  • Timestamps for each cut/action Output as a shooting script with columns: Timestamp | Voiceover | Visual/Shot | Action/Mannerism Notes:

Give Claude maximum context: the reference video URL, product details, target audience pain points, other viral UGC videos in the niche, storytelling frameworks (problem-agitate-solve, before/after, etc.) The more context, the better the script. Claude's reasoning capability is the advantage here. Virality principles to bake in: pattern interrupt hook, social proof, transformation moment, urgency CTA Output: A timestamped shooting script with voiceover, visuals, and creator mannerisms.

Step 2 — Generate the Creator Face (Nano Banana Pro) Tool: Nano Banana Pro (4K — required for realistic skin texture, pores, detail)

Prompt template:

Mix these two faces to create a new attractive face that does not belong to any real person. Use the result as the face of a UGC beauty/lifestyle content creator. Generate at 4K resolution. Hyperrealistic skin texture, natural lighting, creator holding a ring-light mic. Notes:

  • Use 2+ Pinterest/stock face references that match the brand personality (relatable, aspirational, niche-appropriate)
  • Generate with product in hand or near face for the first-frame reference
  • You want skin pores, micro-detail, natural imperfections — this is what makes it not look AI
  • Platform options (cheapest to best):
  • Higgsfield (creator plan) — freemium, works
  • Fal.ai — use coupon pixelriot20hackathon-fal for $20 free credits
  • Google AI Studio — paid, most control
  • Generate 3-5 variants, pick the most realistic one
  • Output: A 4K portrait image of the AI creator holding the product.

Step 3 — Generate the Video (Kling 3.0) Tool: Kling 3.0 (15-second multi-shot model)

Why Kling 3.0:

15-second native output length

Multi-shot: can cut between creator speaking → product closeup → skin transformation → back to creator Expressive: understands creator mannerisms, gestures, personality Has contextual understanding of content platforms (UGC native) Setup:

Input: creator image (Step 2) as the start frame

Input: product image as a reference image (not necessarily start frame — can be mid-video reference) Input: shooting script (Step 1) as the motion/shot prompt Prompt structure:

[First frame: creator holding product, looking at camera, natural lighting] Cut 1 (0-3s): Creator speaks directly to camera with [mannerism], says "[hook line]" Cut 2 (3-7s): Close-up of product being applied / used on skin Cut 3 (7-12s): Creator reaction shot — [emotion/mannerism from script] Cut 4 (12-15s): Creator faces camera, delivers CTA, [mannerism] Notes:

  • Control shots = telling Kling exactly what each cut shows. This is not just prompting — it's directing.
  • Multi-shot = you describe the cuts like a director, not a prompter
  • If confused: watch a Kling 3.0 multi-shot tutorial before attempting
  • Cost: ~$4-5 per 15s clip. Free testing: Fal.ai with coupon pixelriot20hackathon-fal
  • Platforms: Higgsfield, official Kling AI website, Fal.ai
  • Output: A 15-second raw video clip.

Virality Principles (Bake Into Every Step)

Hook = Problem Mirror: Show the audience their own problem in the first 3 seconds. They stop scrolling because they see themselves. Before/After = Hope Loop: After the problem hook, show the transformation. This creates desire. Audio Hook: The whoosh SFX at the cut is a pattern interrupt — keeps watch time up. Creator Mannerisms = Trust: A creator that feels like a real person (laughs nervously, tucks hair, sighs) builds subconscious trust faster than a perfect delivery. Reverb = Room Presence: 10% reverb makes the voice feel like it was recorded in a real space, not a studio. Removes the "AI radio voice" feel. CTA Urgency: Last 2-3 seconds. Direct, specific, low-friction ("link in bio", "tap the link", "DM me").

Branding Note

If this is for a real account/brand (not testing):

  • The creator face should match the brand's target audience persona
  • The creator's style/vibe should be consistent across all UGC videos

Don't mix aesthetics between videos — pick one and lock it Platform / Cost Summary Tool Purpose Cost Gemini 2.5 Pro Script writing Free (AI Studio) Nano Banana Pro Creator face + product images Higgsfield creator plan / Fal.ai ($20 coupon: pixelriot20hackathon-fal) Kling 3.0 Video generation ~$4-5/clip — Higgsfield / Kling AI / Fal.ai ElevenLabs Voice cloning + swap Paid (voice training feature) Play.ht Voice cloning (alt) Freemium Canva Video editing + text Free / Pro Artlist Sound effects Paid (free alternatives exist)

Individual skills in this repo

This repo contains 10 individual skills — each has its own dedicated page.

alecs5am/ralphy

>- Composition and render craft — assembles scenario.json plus asset-manifest.json into a HyperFrames HTML composition and renders the mp4. Owns timing, transitions, captions, audio mix and ducking, the green zone, the render pipeline, and VO sync. Generates no media of its own. USE WHEN the user asks to "compose the video", "render", "preview", "fix the captions", "mix the audio", "tighten the transitions", "final cut", or edits HyperFrames composition code. TRIGGER (EN): "render it", "compose the video", "captions", "audio mix", "transitions", "final cut", "preview the composition".

alecs5am/ralphy

HyperFrames CLI dev loop — `npx hyperframes` for scaffolding (init), validation (lint, inspect), preview, render, and environment troubleshooting (doctor, browser, info, upgrade). Use when running any of these commands or troubleshooting the HyperFrames build/render environment. For asset preprocessing commands (`tts`, `transcribe`, `remove-background`), invoke the `hyperframes-media` skill instead.

alecs5am/ralphy

Asset preprocessing for HyperFrames compositions — text-to-speech narration (Kokoro), audio/video transcription (Whisper), and background removal for transparent overlays (u2net). Use when generating voiceover from text, transcribing speech for captions, removing the background from a video or image to use as a transparent overlay, choosing a TTS voice or whisper model, or chaining these (TTS → transcribe → captions). Each command downloads its own model on first run.

alecs5am/ralphy

Install and wire registry blocks and components into HyperFrames compositions. Use when running hyperframes add, installing a block or component, wiring an installed item into index.html, or working with hyperframes.json. Covers the add command, install locations, block sub-composition wiring, component snippet merging, and registry discovery.

alecs5am/ralphy

Create video compositions, animations, title cards, overlays, captions, voiceovers, audio-reactive visuals, and scene transitions in HyperFrames HTML. Use when asked to build any HTML-based video content, add captions or subtitles synced to audio, generate text-to-speech narration, create audio-reactive animation (beat sync, glow, pulse driven by music), add animated text highlighting (marker sweeps, hand-drawn circles, burst lines, scribble, sketchout), or add transitions between scenes (crossfades, wipes, reveals, shader transitions). Covers composition authoring, timing, media, and the full video production workflow. For dev-loop CLI commands (init, lint, inspect, preview, render) see the hyperframes-cli skill; for asset preprocessing commands (tts, transcribe, remove-background) see the hyperframes-media skill.

alecs5am/ralphy

>- How to make a vivid, character-driven brand POSTER — a single still image, not a video. A generalized overlay on the art-director step: supplies the poster-as-landing architecture (top banner → massive beveled wordmark → chest-up hero photo → scattered vinyl-sticker collage → copy block → thin footer strip), the parameterized slot set (brand, exclamation word, hero, product, palette, original blob glyph, aesthetic register), the model stack (gpt-5.4-image-2 for crisp typography, gemini fallback), and the niche's failure modes. Works for ANY brand and ANY aesthetic the user names — Y2K streetwear, vaporwave, grunge zine, clean DTC — the skill supplies the layout know-how, the user supplies the subject + vibe. USE WHEN the user asks for a poster / flyer / drop poster / "poster-as-landing-page" / brand key-art / album-cover-style still / hype graphic, with no specific existing poster to reproduce. This is a content-niche craft-overlay SKILL (generalized), loaded on top of a poster-format template match — n...

alecs5am/ralphy

>- Compatibility shim for the legacy `/ralphy-remotion` skill name. Use when a saved Claude Code command, old project note, or user request still mentions Ralphy Remotion; redirect the task to HyperFrames and the current Ralphy render playbooks instead of authoring new Remotion code.

alecs5am/ralphy

>- Compatibility shim for the legacy `/remotion-to-hyperframes` skill name. Use when an old project, note, or user request asks to migrate Ralphy Remotion work; translate the intent into the current HyperFrames composition model and follow the editor and HyperFrames playbooks.

alecs5am/ralphy

The full extract → classify → blueprint → de-dup → publish pipeline for a finished `.ralphy/workspaces/<ws>/projects/<id>/`. Reads its `units/*/unit.json` (#069) as the Unit source of truth, then factors the project into ALL FIVE entities — Unit + the four typed blocks Template (structure), Style (look), Recipe (effect/treatment), Asset (reusable media). Applies the recipe-vs-tag split (#082/#083): a candidate is a Recipe only if it carries an extractable artifact (ffmpeg filtergraph / HyperFram

alecs5am/ralphy

>- How to make a realistic AI UGC ad for a product — generalized niche know-how that looks real and doesn't feel AI. A domain overlay on the standard pipeline: supplies the shooting-script shape (timestamp | VO | shot | mannerism), the creator-persona build, the problem-mirror hook, the trust-through-mannerisms rule, and the 9:16 ~15s UGC format defaults. Works for ANY product the user names — their skincare line, a no-name app, a gadget. The skill supplies the know-how; the user supplies the product. USE WHEN the user asks for a UGC ad / creator-style product ad / "make an ad that looks like a real person reviewing my product" / testimonial-style or problem-solution short for TikTok / Reels / Shorts, with no specific existing video to reproduce. This is a niche SKILL (generalized), not a remix TEMPLATE. For "remix this exact ad but swap the product/person", use the remix path (docs/skills-vs-templates.md). For swapping the creator in an existing clip, see /ugc-model-swap.

Skills relacionados

steipete/sag

ElevenLabs text-to-speech with mac-style say UX.

community

steipete/oracle

Oracle CLI second-model review/debug/refactor/design with selected files, dry-run token checks, API or browser engine.

community

steipete/peekaboo

Capture and automate macOS UI with the Peekaboo CLI.

community

mattpocock/design-an-interface

Generate multiple radically different interface designs for a module using parallel sub-agents. Use when user wants to design an API, explore interface options, compare module shapes, or mentions "design it twice".

community

mattpocock/prototype

Build a throwaway prototype to flesh out a design — a runnable terminal app for state/business-logic questions, or several radically different UI variations toggleable from one route.

community

mattpocock/grill-me

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".

community