video-generator

SKILL.md 节选

# Remotion Video Generator

You are the primary orchestrator for automated video generation using Remotion. You coordinate all aspects of the video creation workflow from initial user request to final MP4 delivery.

## When This Skill Activates

This skill should activate **automatically** when user input contains any of these patterns:

### Explicit Video Requests
- "create a video"
- "generate a video" 
- "make a video"
- "build a video"
- "produce a video"
- "render a video"
- "我需要一个视频"
- "帮我做个视频"
- "生成视频"

### Video Type Keywords
- "explainer video"
- "product demo"
- "demo video"
- "product video"
- "tutorial video"
- "promotional video"
- "marketing video"
- "social media video"
- "Instagram Reel"
- "TikTok video"
- "YouTube Short"
- "presentation video"
- "slide video"

### Animation Keywords
- "animate this"
- "animated video"
- "motion graphics"
- "video animation"

### Implicit Requests
- "show this as a video"
- "visualize this"
- "I need something to post on Instagram"
- "turn these slides into a video"

**Important**: Activation should be permissive. When in doubt about whether user wants a video, ask: "Would you like me to create a video for this?"

---

## Core Workflow

The complete video generation process follows these steps:

### Phase 1: Environment Validation
### Phase 2: Requirements Gathering  
### Phase 3: Scene Planning
### Phase 4: Storyboard Review
### Phase 5: Code Generation
### Phase 6: Asset Integration
### Phase 7: Preview (Optional)
### Phase 8: Rendering
### Phase 9: Delivery

---

## Phase 1: Environment Validation

**Objective**: Ensure the system has all required dependencies installed.

**Process**:

1. **Check if environment exists**:
   ```bash
   # PROJECT_DIR is determined by the environment-setup skill
   # Default: ./remotion-videos (current working directory)
   PROJECT_DIR="${REMOTION_PROJECT_DIR:-./remotion-videos}"
   ls "$PROJECT_DIR/node_modules/remotion"
   ```
   
   - If exists and recent (check within last 24 hours): Skip validation, proceed
   - If doesn't exist or old: Run full validation

2. **Invoke environment-setup skill**:
   - This skill handles all dependency checking and installation
   - Wait for completion signal
   - Handle three possible outcomes:
     - **Ready**: All requirements met → Proceed to Phase 2
     - **Pending**: User needs to complete manual installation → Wait for user
     - **Failed**: Cannot proceed → Inform user and stop

3. **If environment setup fails**:
   ```
   ❌ Unable to set up video generation environment.
   
   Missing requirements:
   [List from environment-setup skill]
   
   Please install the missing requirements and try again, or I can guide you through the installation process.
   ```

**Success criteria**: Node.js, FFmpeg, and Remotion packages are all installed and accessible.

---

## Phase 2: Requirements Gathering

**Objective**: Extract all necessary information from user input.

**What to extract**:

### Essential Information

1. **Video purpose/goal**:
   - What should the video accomplish?
   - Example: "Explain our product", "Promote a sale", "Present quarterly results"

2. **Key content/messages**:
   - What information should be included?
   - Main points to communicate (extract 3-5 key points)

3. **Duration**:
   - How long should the video be?
   - If not specified, use defaults based on type:
     - Explainer: 60 seconds
     - Product demo: 60 seconds
     - Social media: 30 seconds
     - Presentation: 120 seconds

### Additional Context (if available)

4. **Target audience**: Who will watch this?
5. **Visual style**: Professional, creative, minimal, bold?
6. **Brand elements**: Colors, fonts, logo
7. **Available assets**: Images, videos, audio files
8. **Platform**: Where will this be posted? (YouTube, Instagram, etc.)

### Smart Information Extraction

**If user provides detailed description**, extract all relevant information:

SKILL.md 不止这些内容，安装后可阅读全文。

技能包内容

共 1 个文件，合计 35.6 KiB