How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

daily-ai-news - Skill - OpenSmartRoute

Name: daily-ai-news
Author: yyh211

Imported from yyh211/claude-meta-skill (daily-ai-news/SKILL.md) via skills.sh. Install upstream with npx skills add yyh211/claude-meta-skill --skill daily-ai-news. Copyright stays with the author.

Daily AI News Briefing

Aggregates the latest AI news from multiple sources and delivers concise summaries with direct links

When to Use This Skill

Activate this skill when the user:

Asks for today's AI news or latest AI developments
Requests a daily AI briefing or updates
Mentions wanting to know what's happening in AI
Asks for AI industry news, trends, or breakthroughs
Wants a summary of recent AI announcements
Says: "给我今天的AI资讯" (Give me today's AI news)
Says: "AI有什么新动态" (What's new in AI)

Workflow Overview

This skill uses a 4-phase workflow to gather, filter, categorize, and present AI news:

Phase 1: Information Gathering
  ├─ Direct website fetching (3-5 major AI news sites)
  └─ Web search with date filters
      ↓
Phase 2: Content Filtering
  ├─ Keep: Last 24-48 hours, major announcements
  └─ Remove: Duplicates, minor updates, old content
      ↓
Phase 3: Categorization
  └─ Organize into 5 categories
      ↓
Phase 4: Output Formatting
  └─ Present with links and structure

Phase 1: Information Gathering

Step 1.1: Fetch from Primary AI News Sources

Use mcp__web_reader__webReader to fetch content from 3-5 major AI news websites:

Recommended Primary Sources (choose 3-5 per session):

VentureBeat AI: https://venturebeat.com/category/ai/
TechCrunch AI: https://techcrunch.com/category/artificial-intelligence/
The Verge AI: https://www.theverge.com/ai-artificial-intelligence
MIT Technology Review AI: https://www.technologyreview.com/topic/artificial-intelligence/
AI News: https://artificialintelligence-news.com/
AI Hub Today: https://ai.hubtoday.app/

Parameters:

return_format: markdown
with_images_summary: false (focus on text content)
timeout: 20 seconds per source

Step 1.2: Execute Web Search Queries

Use WebSearch with date-filtered queries to discover additional news:

Query Template (adjust dates dynamically):

General: "AI news today" OR "artificial intelligence breakthrough" after:[2025-12-23]
Research: "AI research paper" OR "machine learning breakthrough" after:[2025-12-23]
Industry: "AI startup funding" OR "AI company news" after:[2025-12-23]
Products: "AI application launch" OR "new AI tool" after:[2025-12-23]

Best Practices:

Always use current date or yesterday's date in filters
Execute 2-3 queries across different categories
Limit to top 10-15 results per query
Prioritize sources from last 24-48 hours

Step 1.3: Fetch Full Articles

For the top 10-15 most relevant stories from search results:

Extract URLs from search results
Use mcp__web_reader__webReader to fetch full article content
This ensures accurate summarization vs. just using snippets

Phase 2: Content Filtering

Filter Criteria

Keep:

News from last 24-48 hours (preferably today)
Major announcements (product launches, model releases, research breakthroughs)
Industry developments (funding, partnerships, regulations, acquisitions)
Technical advances (new models, techniques, benchmarks)
Significant company updates (OpenAI, Google, Anthropic, etc.)

Remove:

Duplicate stories (same news across multiple sources)
Minor updates or marketing fluff
Content older than 3 days unless highly significant
Non-AI content or tangentially related articles

Deduplication Strategy

When the same story appears in multiple sources:

Keep the most comprehensive version
Note alternative sources in the summary
Prioritize authoritative sources (company blogs > news aggregators)

Phase 3: Categorization

Organize news into 5 categories:

🔥 Major Announcements

Product launches (new AI tools, services, features)
Model releases (GPT updates, Claude features, Gemini capabilities)
Major company announcements (OpenAI, Google, Anthropic, Microsoft, Meta)

🔬 Research & Papers

Academic breakthroughs
New research papers from top conferences
Novel techniques or methodologies
Benchmark achievements

💰 Industry & Business

Funding rounds and investments
Mergers and acquisitions
Partnerships and collaborations
Market trends and analysis

🛠️ Tools & Applications

New AI tools and frameworks
Practical AI applications
Open source releases
Developer resources

🌍 Policy & Ethics

AI regulations and policies
Safety and ethics discussions
Social impact studies
Government initiatives

Phase 4: Output Formatting

Use the following template for consistent output:

# 📰 Daily AI News Briefing

**Date**: [Current Date, e.g., December 24, 2025]
**Sources**: [X] articles from [Y] sources
**Coverage**: Last 24 hours

---

## 🔥 Major Announcements

### [Headline 1]

**Summary**: [One-sentence overview of the news]

**Key Points**:
- [Important detail 1]
- [Important detail 2]
- [Important detail 3]

**Impact**: [Why this matters - 1 sentence]

📅 **Source**: [Publication Name] • [Publication Date]
🔗 **Link**: [URL to original article]

---

### [Headline 2]

[Same format as above]

---

## 🔬 Research & Papers

### [Headline 3]

[Same format as above]

---

## 💰 Industry & Business

### [Headline 4]

[Same format as above]

---

## 🛠️ Tools & Applications

### [Headline 5]

[Same format as above]

---

## 🌍 Policy & Ethics

### [Headline 6]

[Same format as above]

---

## 🎯 Key Takeaways

1. [The biggest news of the day - 1 sentence]
2. [Second most important development - 1 sentence]
3. [An emerging trend worth watching - 1 sentence]

---

**Generated on**: [Timestamp]
**Next update**: Check back tomorrow for the latest AI news

Customization Options

After providing the initial briefing, offer customization:

1. Focus Areas

"Would you like me to focus on specific topics?"

Research papers only
Product launches and tools
Industry news and funding
Specific companies (OpenAI/Google/Anthropic)
Technical tutorials and guides

2. Depth Level

"How detailed should I go?"

Brief: Headlines only (2-3 bullet points per story)
Standard: Summaries + key points (default)
Deep: Include analysis and implications

3. Time Range

"What timeframe?"

Last 24 hours (default)
Last 3 days
Last week
Custom range

4. Format Preference

"How would you like this organized?"

By category (default)
Chronological
By company
By significance

Follow-up Interactions

User: "Tell me more about [story X]"

Action: Use mcp__web_reader__webReader to fetch the full article, provide detailed summary + analysis

User: "What are experts saying about [topic Y]?"

Action: Search for expert opinions, Twitter reactions, analysis pieces

User: "Find similar stories to [story Z]"

Action: Search related topics, provide comparative summary

User: "Only show research papers"

Action: Filter and reorganize output, exclude industry news

Quality Standards

Validation Checklist

All links are valid and accessible
No duplicate stories across categories
All items have timestamps (preferably today)
Summaries are accurate (not hallucinated)
Links lead to original sources, not aggregators
Mix of sources (not all from one publication)
Balance between hype and substance

Error Handling

If webReader fails for a URL → Skip and try next source
If search returns no results → Expand date range or try different query
If too many results → Increase threshold for significance
If content is paywalled → Use available excerpt and note limitation

Examples

Example 1: Basic Request

User: "给我今天的AI资讯"

AI Response: [Executes 4-phase workflow and presents formatted briefing with 5-10 stories across categories]

Example 2: Time-specific Request

User: "What's new in AI this week?"

AI Response: [Adjusts date filters to last 7 days, presents weekly summary]

Example 3: Category-specific Request

User: "Any updates on AI research?"

AI Response: [Focuses on Research & Papers category, includes recent papers and breakthroughs]

Example 4: Follow-up Deep Dive

User: "Tell me more about the GPT-5 announcement"

AI Response: [Fetches full article, provides detailed summary, offers to find expert reactions]

Additional Resources

For comprehensive lists of news sources, search queries, and output templates, refer to:

references/news_sources.md - Complete database of AI news sources
references/search_queries.md - Search query templates by category
references/output_templates.md - Alternative output format templates

Install daily-ai-news and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

{ "ocm": "1", "id": "yyh211-claude-meta-skill-daily-ai-news", "kind": "skill", "name": "daily-ai-news", "description": "Aggregates and summarizes the latest AI news from multiple sources including AI news websites and web search. Provides concise news briefs with direct links to original articles. Activates when user asks for 'today's AI news', 'AI updates', 'latest AI developments', or mentions wanting a 'daily AI briefing'.", "publisher": "yyh211", "version": "1.0.0", "capabilities": { "domains": [ "data_analysis", "math" ], "tags": [ "skill-md", "skills-sh" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Aggregates and summarizes the latest AI news from multiple sources including AI news websites and web search. Provides concise news briefs with direct links to original articles. Activates when user asks for 'today's AI news', 'AI updates', 'latest AI developments', or mentions wanting a 'daily AI briefing'." ], "primary": false, "metadata": { "source": { "provider": "skills.sh", "repository": "https://github.com/yyh211/claude-meta-skill", "path": "daily-ai-news/SKILL.md", "ref": "HEAD", "url": "https://www.skills.sh/yyh211/claude-meta-skill/daily-ai-news", "key": "yyh211/claude-meta-skill/daily-ai-news/SKILL.md" } }, "instructions": "# Daily AI News Briefing\n\n> Aggregates the latest AI news from multiple sources and delivers concise summaries with direct links\n\n## When to Use This Skill\n\nActivate this skill when the user:\n- Asks for today's AI news or latest AI developments\n- Requests a daily AI briefing or updates\n- Mentions wanting to know what's happening in AI\n- Asks for AI industry news, trends, or breakthroughs\n- Wants a summary of recent AI announcements\n- Says: \"给我今天的AI资讯\" (Give me today's AI news)\n- Says: \"AI有什么新动态\" (What's new in AI)\n\n## Workflow Overview\n\nThis skill uses a 4-phase workflow to gather, filter, categorize, and present AI news:\n\n```\nPhase 1: Information Gathering\n ├─ Direct website fetching (3-5 major AI news sites)\n └─ Web search with date filters\n ↓\nPhase 2: Content Filtering\n ├─ Keep: Last 24-48 hours, major announcements\n └─ Remove: Duplicates, minor updates, old content\n ↓\nPhase 3: Categorization\n └─ Organize into 5 categories\n ↓\nPhase 4: Output Formatting\n └─ Present with links and structure\n```\n\n## Phase 1: Information Gathering\n\n### Step 1.1: Fetch from Primary AI News Sources\n\nUse `mcp__web_reader__webReader` to fetch content from 3-5 major AI news websites:\n\n**Recommended Primary Sources** (choose 3-5 per session):\n- VentureBeat AI: https://venturebeat.com/category/ai/\n- TechCrunch AI: https://techcrunch.com/category/artificial-intelligence/\n- The Verge AI: https://www.theverge.com/ai-artificial-intelligence\n- MIT Technology Review AI: https://www.technologyreview.com/topic/artificial-intelligence/\n- AI News: https://artificialintelligence-news.com/\n- AI Hub Today: https://ai.hubtoday.app/\n\n**Parameters**:\n- `return_format`: markdown\n- `with_images_summary`: false (focus on text content)\n- `timeout`: 20 seconds per source\n\n### Step 1.2: Execute Web Search Queries\n\nUse `WebSearch` with date-filtered queries to discover additional news:\n\n**Query Template** (adjust dates dynamically):\n```\nGeneral: \"AI news today\" OR \"artificial intelligence breakthrough\" after:[2025-12-23]\nResearch: \"AI research paper\" OR \"machine learning breakthrough\" after:[2025-12-23]\nIndustry: \"AI startup funding\" OR \"AI company news\" after:[2025-12-23]\nProducts: \"AI application launch\" OR \"new AI tool\" after:[2025-12-23]\n```\n\n**Best Practices**:\n- Always use current date or yesterday's date in filters\n- Execute 2-3 queries across different categories\n- Limit to top 10-15 results per query\n- Prioritize sources from last 24-48 hours\n\n### Step 1.3: Fetch Full Articles\n\nFor the top 10-15 most relevant stories from search results:\n- Extract URLs from search results\n- Use `mcp__web_reader__webReader` to fetch full article content\n- This ensures accurate summarization vs. just using snippets\n\n## Phase 2: Content Filtering\n\n### Filter Criteria\n\n**Keep**:\n- News from last 24-48 hours (preferably today)\n- Major announcements (product launches, model releases, research breakthroughs)\n- Industry developments (funding, partnerships, regulations, acquisitions)\n- Technical advances (new models, techniques, benchmarks)\n- Significant company updates (OpenAI, Google, Anthropic, etc.)\n\n**Remove**:\n- Duplicate stories (same news across multiple sources)\n- Minor updates or marketing fluff\n- Content older than 3 days unless highly significant\n- Non-AI content or tangentially related articles\n\n### Deduplication Strategy\n\nWhen the same story appears in multiple sources:\n- Keep the most comprehensive version\n- Note alternative sources in the summary\n- Prioritize authoritative sources (company blogs > news aggregators)\n\n## Phase 3: Categorization\n\nOrganize news into 5 categories:\n\n### 🔥 Major Announcements\n- Product launches (new AI tools, services, features)\n- Model releases (GPT updates, Claude features, Gemini capabilities)\n- Major company announcements (OpenAI, Google, Anthropic, Microsoft, Meta)\n\n### 🔬 Research & Papers\n- Academic breakthroughs\n- New research papers from top conferences\n- Novel techniques or methodologies\n- Benchmark achievements\n\n### 💰 Industry & Business\n- Funding rounds and investments\n- Mergers and acquisitions\n- Partnerships and collaborations\n- Market trends and analysis\n\n### 🛠️ Tools & Applications\n- New AI tools and frameworks\n- Practical AI applications\n- Open source releases\n- Developer resources\n\n### 🌍 Policy & Ethics\n- AI regulations and policies\n- Safety and ethics discussions\n- Social impact studies\n- Government initiatives\n\n## Phase 4: Output Formatting\n\nUse the following template for consistent output:\n\n```markdown\n# 📰 Daily AI News Briefing\n\n**Date**: [Current Date, e.g., December 24, 2025]\n**Sources**: [X] articles from [Y] sources\n**Coverage**: Last 24 hours\n\n---\n\n## 🔥 Major Announcements\n\n### [Headline 1]\n\n**Summary**: [One-sentence overview of the news]\n\n**Key Points**:\n- [Important detail 1]\n- [Important detail 2]\n- [Important detail 3]\n\n**Impact**: [Why this matters - 1 sentence]\n\n📅 **Source**: [Publication Name] • [Publication Date]\n🔗 **Link**: [URL to original article]\n\n---\n\n### [Headline 2]\n\n[Same format as above]\n\n---\n\n## 🔬 Research & Papers\n\n### [Headline 3]\n\n[Same format as above]\n\n---\n\n## 💰 Industry & Business\n\n### [Headline 4]\n\n[Same format as above]\n\n---\n\n## 🛠️ Tools & Applications\n\n### [Headline 5]\n\n[Same format as above]\n\n---\n\n## 🌍 Policy & Ethics\n\n### [Headline 6]\n\n[Same format as above]\n\n---\n\n## 🎯 Key Takeaways\n\n1. [The biggest news of the day - 1 sentence]\n2. [Second most important development - 1 sentence]\n3. [An emerging trend worth watching - 1 sentence]\n\n---\n\n**Generated on**: [Timestamp]\n**Next update**: Check back tomorrow for the latest AI news\n```\n\n## Customization Options\n\nAfter providing the initial briefing, offer customization:\n\n### 1. Focus Areas\n\"Would you like me to focus on specific topics?\"\n- Research papers only\n- Product launches and tools\n- Industry news and funding\n- Specific companies (OpenAI/Google/Anthropic)\n- Technical tutorials and guides\n\n### 2. Depth Level\n\"How detailed should I go?\"\n- **Brief**: Headlines only (2-3 bullet points per story)\n- **Standard**: Summaries + key points (default)\n- **Deep**: Include analysis and implications\n\n### 3. Time Range\n\"What timeframe?\"\n- Last 24 hours (default)\n- Last 3 days\n- Last week\n- Custom range\n\n### 4. Format Preference\n\"How would you like this organized?\"\n- By category (default)\n- Chronological\n- By company\n- By significance\n\n## Follow-up Interactions\n\n### User: \"Tell me more about [story X]\"\n**Action**: Use `mcp__web_reader__webReader` to fetch the full article, provide detailed summary + analysis\n\n### User: \"What are experts saying about [topic Y]?\"\n**Action**: Search for expert opinions, Twitter reactions, analysis pieces\n\n### User: \"Find similar stories to [story Z]\"\n**Action**: Search related topics, provide comparative summary\n\n### User: \"Only show research papers\"\n**Action**: Filter and reorganize output, exclude industry news\n\n## Quality Standards\n\n### Validation Checklist\n- All links are valid and accessible\n- No duplicate stories across categories\n- All items have timestamps (preferably today)\n- Summaries are accurate (not hallucinated)\n- Links lead to original sources, not aggregators\n- Mix of sources (not all from one publication)\n- Balance between hype and substance\n\n### Error Handling\n- If `webReader` fails for a URL → Skip and try next source\n- If search returns no results → Expand date range or try different query\n- If too many results → Increase threshold for significance\n- If content is paywalled → Use available excerpt and note limitation\n\n## Examples\n\n### Example 1: Basic Request\n\n**User**: \"给我今天的AI资讯\"\n\n**AI Response**:\n[Executes 4-phase workflow and presents formatted briefing with 5-10 stories across categories]\n\n---\n\n### Example 2: Time-specific Request\n\n**User**: \"What's new in AI this week?\"\n\n**AI Response**:\n[Adjusts date filters to last 7 days, presents weekly summary]\n\n---\n\n### Example 3: Category-specific Request\n\n**User**: \"Any updates on AI research?\"\n\n**AI Response**:\n[Focuses on Research & Papers category, includes recent papers and breakthroughs]\n\n---\n\n### Example 4: Follow-up Deep Dive\n\n**User**: \"Tell me more about the GPT-5 announcement\"\n\n**AI Response**:\n[Fetches full article, provides detailed summary, offers to find expert reactions]\n\n## Additional Resources\n\nFor comprehensive lists of news sources, search queries, and output templates, refer to:\n- `references/news_sources.md` - Complete database of AI news sources\n- `references/search_queries.md` - Search query templates by category\n- `references/output_templates.md` - Alternative output format templates", "cost": { "context_tokens": 2138 } }