| name | ai-news-analyzer |
| description | AI news analysis and reporting system that aggregates AI industry news from multiple sources. Use this skill when users ask to analyze AI news, aggregate artificial intelligence trends, track large language model developments, monitor algorithm breakthroughs, or generate comprehensive AI industry reports. This skill fetches AI news via API, enriches topics with web search, analyzes relevance and interest scores, and generates formatted HTML reports with event timelines and product insights. |
| license | MIT |
AI News Analyzer
Overview
This skill analyzes AI industry news and generates comprehensive reports. It aggregates information about artificial intelligence, large language models, algorithms, and AI frontier developments, scoring each topic by relevance and interest value.
Workflow
Phase 1: Fetch AI News
-
Read Configuration: Load api-config.json from the skill directory
- Contains the news API endpoint and authentication details
-
Fetch Raw News: Make API request to retrieve AI news items
- Extract topics, headlines, timestamps, and URLs
- Store in structured format for processing
Phase 2: Enrich with Web Search
For each news topic:
- Extract Key Terms: Identify main keywords from the headline/description
- Web Search: Use WebSearch tool to find:
- Related news coverage
- Background information
- Industry context
- Expert opinions
- Compile Details: Create enriched topic data with:
- Original news item
- Additional context from web search
- Related developments
- Timeline of events
Phase 3: AI Analysis
Analyze each enriched topic using this scoring framework:
Scoring Criteria (Total: 100 points):
-
Interestingness (80 points): How engaging, novel, or surprising is this news?
- Breakthrough innovation: 70-80 points
- Significant development: 60-69 points
- Notable update: 50-59 points
- Routine news: 30-49 points
-
Usefulness (20 points): How practical or actionable is this information?
- Highly actionable insights: 18-20 points
- Valuable context: 15-17 points
- Informative: 10-14 points
- Limited utility: 5-9 points
For each topic, generate:
- Event Timeline: Chronological development of the story
- Product Innovation Details: Technical insights, features, or implications
- Relevance Summary: Why this matters to AI industry
- Comprehensive Score: Total from interestingness + usefulness
Phase 4: Generate HTML Report
Create a comprehensive HTML report using the html-report-generator.js script:
Report Structure:
<!DOCTYPE html>
<html lang="zh-CN">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>AI行业资讯分析报告</title>
<style>
body { font-family: system-ui, -apple-system, sans-serif; }
.topic-card { border-radius: 8px; padding: 20px; margin: 20px 0; }
.excellent { background: linear-gradient(135deg, #667eea 0%, #764ba2 100%); color: white; }
.good { background: linear-gradient(135deg, #f093fb 0%, #f5576c 100%); color: white; }
.normal { background: #f5f5f5; }
.score-badge { display: inline-block; padding: 4px 12px; border-radius: 20px; font-weight: bold; }
</style>
</head>
<body>
<h1>AI行业资讯分析报告</h1>
<p class="report-date">生成时间: [TIMESTAMP]</p>
<div class="topic-card excellent">
<h2>[TOPIC TITLE]</h2>
<div class="score-badge">评分: 85/100</div>
<h3>事件脉络</h3>
<ul class="timeline">
<li>[Timeline item 1]</li>
<li>[Timeline item 2]</li>
</ul>
<h3>产品创意详情</h3>
<p class="product-details">[Product/innovation details]</p>
<h3>综合分析</h3>
<p class="analysis">[Why this matters]</p>
</div>
<div class="summary">
<h2>报告摘要</h2>
<p>本次分析共涵盖 [COUNT] 个AI行业热点话题</p>
<p>优秀话题 (>80分): [EXCELLENT_COUNT] 个</p>
<p>良好话题 (60-80分): [GOOD_COUNT] 个</p>
</div>
</body>
</html>
Scoring Display:
- Excellent (>80 points): Use
.excellent class with purple gradient, prominent display at top
- Good (60-80 points): Use
.good class with pink gradient, display in middle section
- Normal (<60 points): Use
.normal class with gray background, display at bottom
Phase 5: Output
- Save Report: Write HTML to
ai-news-report-[timestamp].html
- Summary: Provide console summary with:
- Total topics analyzed
- Score distribution
- Top 3 highest-scoring topics
- Report file location
File Structure
skills/ai-news-analyzer/
├── SKILL.md # This file
├── api-config.json # News API configuration
├── html-report-generator.js # HTML generation script
└── reports/ # Generated reports directory
Configuration File Template
Create api-config.json:
{
"apiEndpoint": "YOUR_API_ENDPOINT_HERE",
"apiKey": "YOUR_API_KEY_HERE",
"sources": ["techcrunch", "venturebeat", "arxiv", "nature"],
"categories": ["artificial-intelligence", "machine-learning", "deep-learning", "llm"],
"maxTopics": 20
}
Usage Examples
Example 1: Generate daily AI news report
"Generate today's AI news analysis report"
Example 2: Analyze specific AI topics
"Analyze recent AI developments and create a report"
Example 3: Weekly AI industry summary
"Create a comprehensive weekly AI industry analysis"
Error Handling
- API Failure: If API call fails, ask user to verify endpoint and credentials in
api-config.json
- No Results: If no news found, inform user and suggest checking API configuration
- Search Rate Limits: If web search hits limits, pause and retry after delay
- HTML Generation Error: If report generation fails, provide raw JSON output as fallback
Limitations
- Requires user to provide valid news API endpoint and credentials
- Web search is limited to publicly available information
- Scoring is subjective and based on AI analysis
- Report quality depends on source article quality
- Real-time news may have latency
Best Practices
- API Security: Never commit
api-config.json with real credentials. Use environment variables
- Rate Limiting: Implement delays between web searches to avoid hitting rate limits
- Source Diversity: Aggregate from multiple sources for balanced coverage
- Date Filtering: Always include date filters to avoid stale news
- User Feedback: Allow users to adjust scoring thresholds or exclude sources
Output Quality Standards
All reports must include:
Testing Checklist
Before considering the skill complete:
