Turn a Brand URL Into Phone-Shot AI UGC Videos With Kling 3 and Sora 2 — workflow skill for ads & UGC
For AI agents: find the right skill in one call —
curl -s "https://varg.ai/how-to/api/search?q=<what the person wants>"returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://varg.ai/how-to/llms.txt. Every page has a markdown twin: add.mdto its path. Before generating anything, show the person the options on the how-to page. From a terminal:curl -fsSL https://varg.ai/install | shinstalls the vargai CLI (https://varg.ai/how-to/cli.md):vargai searchandvargai install <slug>with no key, every varg model aftervargai login.
Brand URL to anti-polish UGC video: persona research, Nano Banana first frame, Sora 2/Kling 3 A-roll, Kling 3 B-roll, captions, virality score.
- Type: Workflow · Ads & UGC
- Slug:
turn-a-brand-url-into-phone-shot-ai-ugc-videos-kling-sora - Status: live
- Publisher: TheMattBerman (community)
- Source: https://github.com/TheMattBerman/scrollclaw
- Works in: Claude Code
What it does
A suite of sub-skills that takes a brand name and URL through brand setup, persona research mined from real reviews, a locked AI creator, a script mapped to one of six formats (talking head, hook face + demo, podcast clip, wall of text, visual and hybrid transformation), an iPhone-style first frame, A-roll, B-roll, grain-and-grade post-production, captions and a 7-criteria virality gate (publish at 70+). A-roll runs on Sora 2 with Kling 3 as the fallback, B-roll on Kling 3 seeded from an A-roll frame. Needs FAL_KEY and REPLICATE_API_TOKEN (OpenRouter recommended, ElevenLabs optional), Python 3 with PIL and ffmpeg; the README puts a video at $5-$50 in API costs.
Use when
- Make a UGC video or run a UGC campaign for a brand
- Write a UGC script from real customer review language
- Make a talking-head product review that looks shot on a phone
- Add environment-matched B-roll and native-style captions to an AI UGC clip
- Score a UGC video for virality before publishing
Install
Not published as a standalone skill yet. Source: https://github.com/TheMattBerman/scrollclaw
More ads & UGC skills
- Ecommerce video ad skills for Claude Code and AI agents: Product videos that sell on TikTok, TikTok Shop, Reels and Meta: angle, hook, format, script, rules, variants.
- Turn a product into scroll-stopping video hooks: The first 3 seconds of a product video: frame, line and text that agree, from a 70-line hook bank.
- Open-source alternative to Arcads: AI UGC ads: AI UGC ads with an AI actor holding your real product: face, voice and label locked, labelled.
- Turn one winning ad into a creative testing plan: How many ads to test, what to change, when to judge, kill or scale — concepts Meta sees as different.
- Turn a product into a dropshipping guide reel: The faceless "comment GUIDE" reel: product hook, a marker pointing at cost and price on a laptop, a skyline ending.
- Turn product photos into a viral product hook: Find a hook that already sells, then remake it with your product as phone or security-cam footage on Seedance 2.5.
Run the generation on varg
varg is one API key for the models these skills drive — Seedance 2.5, Kling 4, Veo 3.1, Wan 3, Nano Banana 2.1, GPT Image 2.5, ElevenLabs v4 and 200+ more — billed in credits, with a cost estimate before anything runs.
- Any agent with a terminal:
curl -fsSL https://varg.ai/install | sh(Windows:irm https://varg.ai/install.ps1 | iex), thenvargai loginand every model is a command —vargai generate_video …,vargai estimate_cost …(https://varg.ai/how-to/cli/) - Claude Code:
claude plugin marketplace add vargHQ/skills && claude plugin install varg@varg(skill + hosted MCP) - Codex, Cursor, Gemini CLI, OpenClaw:
npx skills add vargHQ/skills --skill varg-ai -a <agent> -y - Grok and any agent that reads SKILL.md:
npx skills add vargHQ/skills --skill varg-ai, or point it at the skill's SKILL.md - Claude.ai, ChatGPT, any MCP client: connect
https://mcp.varg.ai/mcp - API key: https://varg.ai/dashboard/api · Model specs: https://varg.ai/models/llms.txt