Video Generation Skill

Evolink Video — AI Video Generation Skill for OpenClaw

---

What Is This?

An OpenClaw skill powered by EvoLink. Install one skill, and your AI agent gains the ability to generate videos with 37 models from the world's leading providers — all through a single API.

This is the video-focused view of evolink-media. Install the full skill for image and music generation too.

🚀 Try All Models Now →

More skills coming soon.

---

Installation

Quick Install (Recommended)

openclaw skills add https://github.com/EvoLinkAI/video-generation-skill-for-openclaw

Done. The skill is ready for your agent.

Manual Install

git clone https://github.com/EvoLinkAI/video-generation-skill-for-openclaw.git
cd video-generation-skill-for-openclaw
openclaw skills add .

---

Get API Key

1. Sign up at evolink.ai
2. Go to Dashboard → API Keys
3. Create a new Key
4. Set the environment variable:

export EVOLINK_API_KEY=your_key_here

Or just tell your OpenClaw agent: "Set my EvoLink API key to ..." — it'll handle the rest.

---

Video Generation

Generate AI videos through natural conversation with your OpenClaw agent.

Capabilities

• Text-to-Video — Describe a scene, get a video
• Image-to-Video — Provide reference images to guide output
• First-Last-Frame — Define start and end frames for precise control
• Video Editing — Edit and transform existing videos
• Auto Audio — Sync voice, sound effects, and background music
• Multi-Resolution — 480p, 720p, 1080p, 4K
• Flexible Duration — 3–15 seconds (model dependent)
• Aspect Ratios — 16:9, 9:16, 1:1, 4:3, 3:4, 21:9

Usage Examples

Just talk to your agent:

"Generate a 5-second video of a cat playing piano"

"Create a cinematic ocean sunset shot, 1080p, 16:9"

"Use this image as reference and generate an 8-second animated video"

"Edit this video to add a dreamy slow-motion effect"

The agent will guide you through missing details and handle generation.

Models (37)

Top Picks

Stable Models (26)

seedance-1.5-pro, seedance-2.0, doubao-seedance-1.0-pro-fast, sora-2-preview, kling-o3-text-to-video, kling-o3-image-to-video, kling-o3-reference-to-video, kling-o3-video-edit, kling-v3-text-to-video, kling-v3-image-to-video, kling-o1-image-to-video, kling-o1-video-edit, kling-o1-video-edit-fast, kling-custom-element, veo-3.1-generate-preview, veo-3.1-fast-generate-preview, MiniMax-Hailuo-2.3, MiniMax-Hailuo-2.3-Fast, MiniMax-Hailuo-02, wan2.5-t2v-preview, wan2.5-i2v-preview, wan2.5-text-to-video, wan2.5-image-to-video, wan2.6-text-to-video, wan2.6-image-to-video, wan2.6-reference-video

Beta Models (11)

sora-2, sora-2-pro, sora-2-beta-max, sora-character, veo3.1-pro, veo3.1-fast, veo3.1-fast-extend, veo3, veo3-fast, grok-imagine-text-to-video, grok-imagine-image-to-video

MCP Tools

Parameters

---

Setup — MCP Server Bridge

For the full tool experience, bridge the MCP server @evolinkai/evolink-media (npm):

Via mcporter:

mcporter call --stdio "npx -y @evolinkai/evolink-media@latest" list_models

Or add to mcporter config:

{
  "evolink-media": {
    "transport": "stdio",
    "command": "npx",
    "args": ["-y", "@evolinkai/evolink-media@latest"],
    "env": { "EVOLINK_API_KEY": "your-key-here" }
  }
}

Direct install (Claude Code):

claude mcp add evolink-media -e EVOLINK_API_KEY=your-key -- npx -y @evolinkai/evolink-media@latest

---

File Upload

For image-to-video generation, upload reference images first:

1. Call upload_file with file_path, base64_data, or file_url → get file_url (synchronous)
2. Use that file_url as image_urls input for generate_video

Supported: Images (JPEG, PNG, GIF, WebP). Max 100MB. Files expire after 72 hours. Quota: 100 files (default) / 500 (VIP).

---

Workflow

1. Upload reference images if needed (via upload_file)
2. Call generate_video → get task_id
3. Poll check_task every 10–15s until completed
4. Download result URLs (expire in 24h)

---

Script Reference

The skill includes scripts/evolink-video-gen.sh for direct command-line usage:

# Text-to-video
./scripts/evolink-video-gen.sh "Serene mountain dawn landscape" --duration 5 --quality 720p

# With reference image
./scripts/evolink-video-gen.sh "Gentle ocean waves" --image "https://example.com/beach.jpg" --duration 8 --quality 1080p

# Vertical (social media)
./scripts/evolink-video-gen.sh "Dancing particles" --aspect-ratio 9:16 --duration 4 --quality 720p

# Choose a specific model
./scripts/evolink-video-gen.sh "Cinematic car chase" --model sora-2-preview --duration 10 --quality 1080p

# No audio
./scripts/evolink-video-gen.sh "Abstract art animation" --duration 6 --quality 720p --no-audio

API Reference

Full API documentation: references/api-params.md

---

File Structure

.
├── README.md