初始化

This commit is contained in:
TCubic
2026-02-04 18:35:05 +08:00
commit 411365fd64
12 changed files with 1798 additions and 0 deletions

View File

@@ -0,0 +1,97 @@
# 🎬 Viral Topic Generator for Google Antigravity
## 📖 Overview
This Agent Skill analyzes video content to extract successful patterns and generates viral science communication topics. Compatible with Google Antigravity and other AI coding assistants supporting the Agent Skills standard.
## 🚀 Installation & Usage
### Installation
This skill is located in `.agent/skills/viral-topic-generator/` for workspace-specific use, or can be installed globally to `~/.gemini/antigravity/skills/viral-topic-generator/`.
### Usage in Antigravity
```bash
# Direct invocation
/viral-topic-generator [video-url]
# Natural language trigger
"Analyze this video and generate similar topics: [video-url]"
```
## 🎯 Compatibility
### Supported Platforms
- ✅ Google Antigravity
- ✅ Claude Code
- ✅ Cursor
- ✅ GitHub Copilot
- ✅ Gemini CLI
- ✅ Other Agent Skills-compatible tools
### Video Platforms
- ✅ YouTube
- ✅ Bilibili
- ✅ TikTok/Douyin
- ✅ Instagram Reels
- ✅ LinkedIn Video
- ✅ Other mainstream platforms
## 📊 What This Skill Does
1. **🔍 Video Analysis**
- Extracts script structure patterns
- Analyzes narrative rhythm and pacing
- Identifies educational logic frameworks
2. **🎬 Content Generation**
- Creates 5 original topic concepts
- Provides detailed content outlines
- Includes viral element strategies
3. **📈 Strategic Insights**
- Performance predictions
- Platform adaptation suggestions
- Production recommendations
## 💡 Best Practices
### Choosing Reference Videos
- Select high-performing content (10K+ views recommended)
- Focus on educational/science communication content
- Choose videos with clear structure and good engagement
### Using Generated Topics
- Adapt topics to your expertise area
- Customize for your target platform
- Maintain authenticity while applying proven structures
## 🔄 Follow-up Prompts
After generating topics, try:
- "Expand topic #2 into a full script"
- "Adapt these topics for TikTok format"
- "How can I test these topics' viral potential?"
- "Create a content calendar using these topics"
## 📁 File Structure
```
.agent/skills/viral-topic-generator/
├── SKILL.md # Main skill definition
└── README.md # This usage guide
```
## 🌟 Agent Skills Standard
This skill follows the open Agent Skills standard, making it portable across different AI development environments. The SKILL.md format uses YAML frontmatter for metadata and markdown for instructions.
## 🤝 Contributing
This skill can be enhanced with:
- Platform-specific analysis features
- Multi-language support improvements
- Advanced viral prediction algorithms
- Integration with analytics APIs
---
Ready to create viral science content? Load this skill and start analyzing successful videos to fuel your content strategy!

View File

@@ -0,0 +1,127 @@
---
name: viral-topic-generator
description: Analyzes reference videos to extract script structure, narrative rhythm, and educational logic, then generates 5 viral science communication video topics in similar style. Use when users want to create content based on successful video examples.
argument-hint: [video-url]
user-invocable: true
---
# 🎬 Viral Topic Generator - Video Content Analysis & Planning
Analyze reference video: $ARGUMENTS and generate strategic content plans.
## 📋 Analysis Workflow
### Step 1: Video Information Extraction
1. **Basic Video Data**
- Title, description, duration
- Platform and performance metrics
- Target audience positioning
### Step 2: Deep Content Analysis
2. **Script Structure Analysis**
- Hook design (first 3 seconds)
- Main content architecture (problem → explanation → examples → conclusion)
- Ending design (summary + engagement triggers)
- Transition techniques and pacing control
3. **Narrative Rhythm Analysis**
- Information density distribution
- Emotional arc and peaks
- Key moment timing allocation
- Visual element synchronization
4. **Educational Logic Analysis**
- Knowledge hierarchy structure
- Complex concept simplification methods
- Analogies and metaphors usage
- Interactive question design
- Memory reinforcement techniques
### Step 3: Style Pattern Extraction
5. **Expression Style Features**
- Language style (formal/casual/humorous/technical)
- Technical terminology density
- Interaction preferences
- Visual presentation characteristics
### Step 4: Topic Generation
6. **Generate 5 New Topics**
For each topic provide:
**📺 Topic Title**: Eye-catching headline
**🎯 Target Audience**: Precise audience profile
**⏱️ Estimated Duration**: Recommended video length
**📖 Content Outline**:
- Hook opening (0-3 seconds)
- Problem introduction (3-15 seconds)
- Core content (15 seconds - 80% of duration)
- Summary & conclusion (last 20%)
**🔥 Viral Elements**:
- Controversy/curiosity triggers
- Engagement strategies
- Memory hooks design
- Shareability motivations
**📊 Performance Prediction**: Data expectations based on reference style
**🎨 Production Recommendations**:
- Key visual elements
- Pacing control points
- Interaction design suggestions
## 📈 Output Format
```
# 📺 Viral Topic Analysis Report
## 🔍 Reference Video Analysis
[Detailed analysis content]
## 🎬 Generated Topics (5 Total)
### Topic 1: [Title]
**Target Audience**: xxx
**Estimated Duration**: xxx
**Content Outline**:
- Hook opening: xxx
- Problem introduction: xxx
- Core content: xxx
- Summary & conclusion: xxx
**Viral Elements**: xxx
**Performance Prediction**: xxx
**Production Recommendations**: xxx
---
[Other 4 topics in same format]
```
## 🎯 Core Principles
1. **Preserve Success DNA**: Inherit core success factors from reference video
2. **Differentiated Innovation**: Innovate content while maintaining proven style elements
3. **Audience-Centric**: Ensure topics match target audience interests and comprehension level
4. **Actionable Guidance**: Provide specific, executable production instructions
5. **Viral Potential**: Each topic must contain elements for viral spread
## 🛠️ Tool Requirements
For optimal performance, this skill works best with agents that can:
- Fetch web content from video URLs
- Search for additional context and data
- Read and analyze video transcripts or descriptions
- Generate structured content output
## 📝 Usage Notes
- Supports major video platforms: YouTube, Bilibili, TikTok, etc.
- Works with both English and Chinese content
- Optimized for science communication and educational content
- Generates platform-agnostic content that can be adapted for different social media
Begin analysis of reference video and strategic topic generation...