DEV Community: ZeroLu

Let Your AI Agent Generate Images and Videos Automatically - Open Source Skill

ZeroLu — Fri, 06 Mar 2026 14:47:45 +0000

AI coding tools are getting more powerful every day. Tools like Claude Code, Codex, and Cursor can write code, debug, and even architect entire applications.

But there's one thing they still can't do: generate images and videos.

That's why I open-sourced Ultimate AI Media Generator Skill.

What Does It Do?

It gives your AI Agent the ability to call top-tier image and video generation models:

Image Models: Nano Banana 2, Nano Banana Pro
Video Models: Sora 2, Seedance 1.0/2.0, Kling 2.6, Veo 3.1

Quick Start

Install with one command:

npx skills add ZeroLu/Ultimate-AI-Media-Generator-Skill --all

Then configure your API Key following the README instructions.

Usage

Just tell your AI Agent: "Generate a cyberpunk city night scene, 16:9, 1K resolution"

And it will return the generated image URL in seconds.

Supported Platforms

OpenClaw, Claude Code, Codex, Cursor, Antigravity, and any skill-enabled AI platform.

Real Use Cases

UI Mockups: Generate design assets while coding
Blog Illustrations: Auto-generate images for articles
Marketing Materials: Batch generate product visuals
Video Creation: Create short video clips

GitHub

https://github.com/ZeroLu/Ultimate-AI-Media-Generator-Skill

MIT License. Stars and contributions welcome!

Curated Seedance 2.0 Prompts for Cinematic, Anime, UGC, and Viral AI Videos

ZeroLu — Thu, 26 Feb 2026 06:55:29 +0000

With Seedance 2.0 gaining massive attention for its high-fidelity video generation capabilities, many creators outside China struggle to find quality prompts and resources.

I've compiled a curated collection of prompts from top prompt engineers on X/Twitter, covering various styles and use cases.

🎬 What's Inside

Category	Description
Cinematic Film Styles	Hollywood racing, Denis Villeneuve epic, Wong Kar-wai art cinema
Advertising & Commercial	Product showcases, brand storytelling, motion graphics
Social Media & Viral Memes	Attention-grabbing content, comedic surrealism
UGC Style	Mockumentary, vlog aesthetics, realistic footage
Anime & Animation	Character consistency, dynamic motion, tournament battles
Short-form Drama	Mini-series, emotional scenes, viral content
Visual Effects	Surrealism, experimental styles, epic visuals

📝 Prompt Examples

1. Hollywood Professional Racing Movie (Le Mans Style)

Style: Hollywood Professional Racing Movie (Le Mans Style), Cinematic Night, Rain, High Stakes Sport.
Duration: 15s.

[00-05s] Shot 1: The Veteran (Interior/Close-up).
Rain lashes the windshield of a high-tech race car on a track. The Veteran driver (in helmet) looks over, calm and focused. Dashboard lights reflect on his visor.
Dialogue Cue: He gives a subtle nod and mouths "Let's go."

[05-10s] Shot 2: The Challenger (Interior/Close-up).
Cut to the rival car next to him. The younger driver grips the wheel tight, breathing heavily. Eyes wide with adrenaline.
Dialogue Cue: He whispers "Focus" to himself.

[10-15s] Shot 3: The Green Light (Wide Action).
The starting lights turn Green. Both cars accelerate in perfect sync on the wet asphalt. Water sprays massively into the camera lens. Motion blur turns the stadium lights into long streaks of color.

🎬 Watch Video

2. Wong Kar-wai Film Style (Rainy Phone Booth Scene)

[Film Style]: 90s Hong Kong Art Cinema style, retro film feel, high ISO grain, ambiguous yellow-green tint, frame stepping effect, melancholic atmosphere.

[Core Dialogue (for emotion control)]: "If memories were canned food, I hope they never expire."

[Video Duration]: 10 seconds

[00:00-00:04] Shot 1: Through the Glass Peeping.
Scene: A rain-covered red public telephone booth.
Character: A man (or woman) in a khaki trench coat holding the receiver tightly, not speaking, just listening.
Emotional Performance: Through the glass refraction, see his/her eyes hollow yet deeply emotional.

[00:04-00:07] Shot 2: Extreme Close-up & Micro-expression.
Scene: Focus on the character's lips and half face.
Action: He/She whispers softly into the receiver. Lips tremble slightly.
Lighting: Street neon bokeh flows across his face, bright and dim alternately.

[00:07-00:10] Shot 3: Signature Slow-shutter Drag Shadow.
Scene: Character hangs up phone, turns around and walks into the rainy crowd.
Visual Effect: Using frame stepping effect (stop-motion feel), the character's back becomes blurred with trailing shadows.

🎬 Watch Video

3. Giant Orange Cat Meme Style

【Style】Mockumentary, mobile Vlog perspective, hyperrealistic CG combined with real scenes, 8K quality, perfect fur physics simulation.
【Duration】15 seconds
【Scene】Hongya Cave in Chongqing or a busy overpass intersection.

[00:00-00:05] Shot 1: Visual spectacle (The Reveal).
The scene shows a bustling city street. The camera lifts up to reveal a **Godzilla-sized orange tabby cat** stuck between two skyscrapers.
Action: The giant cat is stuck because it's too fat, waving its huge paws with a pitiful expression, trying to pull itself out.
Detail: Cat fur is clearly visible in the sunlight, huge paw pads pressing against glass curtain walls.

[00:05-00:10] Shot 2: Absurd interaction (The Interaction).
The camera switches to ground-level perspective. The giant cat lowers its head, curiously sniffing a bus waiting at a red light.
Action: The bus driver calmly reaches out and pets the giant cat's nose. The cat sneezes, instantly blowing away roadside leaves.

[00:10-00:15] Shot 3: Memetic ending (The Punchline).
The giant cat finally squeezes past the buildings and sits down on a cross-river bridge, blocking the entire evening rush hour traffic.

🎬 Watch Video

4. Surrealistic Documentary Style (Mirror Glitch)

【Style】Mockumentary (Vlog Style), hyperrealism, fixed-camera real-shot feel, natural lighting, with a slight suspenseful comedy tone.
【Duration】15 seconds
【Main Character】An ordinary young beautiful woman, in front of the bathroom sink at home.

[00:00-00:06] Shot 1: Daily setup (Normalcy).
Scene: In front of a regular bathroom mirror.
Action: The protagonist is brushing her teeth, mouth full of foam. She makes various funny faces at the mirror while brushing.
Key detail: At this point, the reflection in the mirror is completely normal, movements synchronized.

[00:06-00:11] Shot 2: BUG appears (The Glitch).
Action: After brushing teeth, the protagonist lowers her head to spit out foam, then turns around to leave the bathroom.
High-impact moment: Just as the protagonist's real body has turned and left the mirror frame, the "reflection" in the mirror **doesn't move**! That "reflection" still maintains the tooth-brushing pose, even mischievously raising eyebrows at the camera with a bad smile.

[00:11-00:15] Shot 3: Comedic callback (The Punchline).
Action: The protagonist, who has already walked to the door, seems to sense something is wrong, suddenly turning back to look at the mirror.
Result: The mirror has now completely returned to normal, only reflecting the opposite wall. The protagonist scratches her head in confusion.

🎬 Watch Video

5. Chinese Mini-Drama Style (Rainy Night Emotional Scene)

【Style】Popular Chinese web drama style (Mini-Drama Style), extreme fast-cut rhythm, high attractiveness filter, emotional outburst, romantic and heart-wrenching rainy night.
【Duration】15 seconds
【Characters】Deeply affectionate tycoon male lead (black coat, wet hair, red-rimmed eyes) VS stubbornly broken-hearted female lead (white dress, face full of tears).

[00:00-00:05] Shot 1: Rapid cut combination.
Rainy street. Female lead decisively turns to leave (back view). Male lead rushes up and grabs her wrist (close-up). Female lead suddenly turns back, eyes showing pain filled with love and hate.
【Dialogue lip-sync guidance】Female lead cries out: "Let go! We're done!"

[00:05-00:10] Shot 2: Truth explosion (Intense Close-ups).
Male lead refuses to let go, rainwater streaming down both their faces. Male lead urgently pulls out a ring (or a document) from his chest, raises it in front of her, fingers trembling.
【Dialogue lip-sync guidance】Male lead shouts: "Look carefully! I never deceived you!"

[00:10-00:15] Shot 3: Emotional dam burst (Climax).
At the moment the female lead sees the object in his hand, her pupils shake (extreme close-up), covers her mouth, defenses collapse. The next second, the male lead suddenly pulls her into his arms and holds her tightly.

🎬 Watch Video

6. Van Gogh Post-Impressionism Style Animation

【Style】Van Gogh Post-Impressionism oil painting, thick paint texture (Heavy Impasto), signature swirling brushstrokes, dreamy feel, high saturation blue-yellow contrast.
【Duration】15 second animation

[Visual Content] This is a completely dynamic world made of thick oil paint.
Sky: In the deep blue night sky, huge yellow celestial bodies and crescent moon are surrounded by radiating short brushstrokes, nebulae swirling wildly in the air like rushing rivers.
Foreground: On the left is a huge cypress tree, twisted like burning black flames, shooting up to the sky.
Background: A sleeping town in the valley, windows of houses emit warm, circularly painted yellow light.

🎬 Watch Video

🔗 Resources

GitHub Repo: ZeroLu/awesome-seedance - Full collection with video examples
Seedance 2.0 Official: Coming soon to Cyberbara

🤝 Contributing

Contributions welcome! If you have awesome Seedance 2.0 prompts, submit a Pull Request.

Happy creating! 🎬✨

HOW TO REMOVE WATERMARKS FROM SORA AI VIDEOS - A COMPLETE GUIDE

ZeroLu — Mon, 03 Nov 2025 05:49:05 +0000

HOW TO REMOVE WATERMARKS FROM SORA AI VIDEOS - A COMPLETE GUIDE

A simple, step-by-step guide to using TheSoraWatermarkRemover.com, a free online tool to get clean, high-quality videos from OpenAI's Sora.

➡️ Visit the Free Sora Watermark Remover Website

The Problem: Annoying Dynamic Watermarks

Since Sora's launch, its stunning AI-generated videos have gone viral on platforms like TikTok. However, a major frustration for creators is the persistent dynamic watermark.

OpenAI has made it particularly challenging to get a clean version: screen recording is often blocked on the app, and every video on the web has this moving watermark embedded. Because it's dynamic, traditional removal methods often fail or leave ugly artifacts.

The Solution: A Simple Online Tool

Fortunately, there's a simple and effective solution: TheSoraWatermarkRemover.com.

This free online tool allows you to remove the Sora watermark in seconds, with 1 free uses upon login. No registration, no payment, no software installation needed.

📖 Step-by-Step Guide

Follow these three simple steps to download your Sora video without the watermark.

Step 1: Copy Your Sora Video Link

You need the direct link to the Sora video. Here is how to get it from both the web and the mobile app.

From the Web:
Click the three dots (...) on the video and then select "Copy link".

From the App:
Tap the three dots (...) or the share icon (paper airplane), and then choose "Copy link".

Step 2: Paste the Link and Process

Go to thesorawatermarkremover.com, paste the video link you just copied into the input box, and click the button to start.

Step 3: Download Your Watermark-Free Video

The video is typically processed in under 3 seconds. Once it's ready, simply click the download button to save the clean video to your device.

✨ The Final Result

The final video is perfectly clean with no visible watermark or processing flaws. It's ready for you to share or use in your projects!

⭐ Key Features

💧 High-Quality Removal: Advanced algorithms cleanly remove the dynamic watermark without degrading video quality.
🆓 Free to Use: Get 1 free watermark removals upon login.
🚀 Incredibly Fast: Processes most videos in under 3 seconds.
🚫 No Login Required: No need to create an account or provide any personal information.
💻 100% Online: Works directly in your browser on any device (desktop or mobile). No software to install.

⚠️ Disclaimer

This guide and the associated website are provided for educational and research purposes only. The ability to remove a watermark does not grant you ownership or copyright of the video content.

Please respect OpenAI's Terms of Service and the intellectual property rights of the content. Users are solely responsible for their actions and any potential legal consequences. We do not endorse the use of this tool for any illegal or unethical activities.