News Aggregator: Multi-Source Intelligence for Your Agents

Where this fits

You're building a research agent that needs to answer "What's the latest on the EU AI Act?" or "Is this startup getting press momentum?" A single news source gives you one perspective, missed stories, and stale results. The news-aggregator skill pulls from Google News, Tavily, and Exa simultaneously, cross-references publication velocity, and surfaces emerging stories—so your agent can deliver a synthesized, citation-ready briefing without you stitching together three separate API calls.

Why agents benefit

• Cross-source validation: Detect consensus or contradictions by comparing headlines and timestamps across three independent providers.
• Publication velocity tracking: Identify when a topic is accelerating (multiple new articles per hour) vs. steady—useful for spotting breaking news or fading trends.
• Emerging story detection: Exa's semantic search and Tavily's freshness filters catch stories that haven't hit mainstream Google News yet.
• Citation-ready output: Every result includes source URLs and publish dates, so your agent can generate references without post-processing.
• Schema-driven API calls: The skill enforces sandbase_describe_tool before each call, reducing hallucinated arguments and API errors.

Practical scenarios

Competitive intelligence for product launches

Your agent monitors a competitor's funding announcements. Instead of querying one endpoint, it aggregates Crunchbase-style news from Google News, Tavily's web search, and Exa's semantic search. It then ranks results by recency and source authority, flagging a spike in coverage that might indicate a press release.

Journalist's research assistant

A reporter asks for "all major coverage of the water crisis in the last 48 hours." The skill pulls articles, deduplicates overlapping stories, and summarizes key facts with links. The agent can also identify which outlets broke the story first by comparing timestamps across sources.

Real-time event monitoring

For a trading bot, you need to react to market-moving news. The skill's velocity metric—how many articles per hour per source—acts as a signal for unusual activity. The agent can trigger alerts when velocity exceeds a threshold, then provide a concise digest with citations.

Add it to your agent workflow

Install the skill directly:

npx skills add sandbaseai/sandbase-skills --skill news-aggregator

Then in your agent's tool-calling loop, follow the README's pattern:

# Pseudocode for an agent step
await sandbase_describe_tool("tavily_search")
result = await sandbase_call_tool(
    tool_name="tavily_search",
    arguments={"query": "EU AI Act latest", "days": 3}
)
# Repeat for google_news_bulk_articles and exa_search
# Merge results and compute velocity

For a deeper look at how this skill fits into the broader agent-skill ecosystem, see the skill page.

Compare before adopting

Before wiring this in, check three things: quality signals—does the skill handle deduplication and source credibility ranking, or just dump raw results? Maintenance freshness—the repository was last updated recently (68 stars, active author), but verify the API schemas haven't changed. Alternatives—a simple tavily_search alone might suffice if you don't need multi-source velocity. Workflow fit—does your agent already have a news tool? If so, weigh the complexity of three API calls vs. the added intelligence.

Why it is worth tracking

This skill is a solid example of a research-grade capability that goes beyond a single API wrapper. The author (sandbaseai) maintains a set of skills with consistent patterns, and the 68-star count suggests community validation. Track it if you're building agents that need timely, cross-verified information—especially in fast-moving domains like finance or tech. Reevaluate when you need deeper source analysis (e.g., sentiment or entity extraction) that this skill doesn't currently provide.