🔑 Key Concepts
- Description Field: The Agent decides when to use this tool based on the description you provide. Be clear and precise about what the tool is meant to search (e.g., “Use this tool to fetch the latest news articles related to financial markets”).
- Topics: Define whether the search should be general or restricted to news.
- Maximum Days: Set a timeframe filter to restrict results to recent content (e.g., last 7, 30, or 90 days).
- Mode:
- Basic: Standard web search with less filtering.
- Advanced: More precise search logic, with better contextual alignment to the Agent’s goal.
- Results: The maximum number of results the tool should return (e.g., top 5, top 10).
⚙️ Configuration Fields
-
Name – Give the tool a meaningful name that reflects its purpose.
Example:Latest Tech News Search. -
Description – Write a clear instruction for the Agent.
Example: -
Topic – Choose between:
General→ Broader search results from the web.News→ Results restricted to news sources.
-
Maximum Days – Define the time range for content freshness.
- Example:
7→ Only last week’s content.
- Example:
-
Mode –
Basic→ Quick, broad searches.Advanced→ More structured and reliable searches.
-
Results – The number of search results to be returned (1–20).
- Example:
10→ Top 10 results.
- Example:
📘 Example Setup
- Name:
Crypto News Search - Description:
Use this tool to fetch the latest cryptocurrency news from reliable sources. - Topic:
News - Maximum Days:
30 - Mode:
Advanced - Results:
5
💡 Best Practices
- Keep the description short but precise so the Agent knows exactly when to trigger this tool.
- Use News mode for contexts requiring trust and recency (finance, politics, technology updates).
- Use General mode when open web context is sufficient (definitions, product details, how-to guides).
- Set Maximum Days thoughtfully – shorter ranges for fast-moving industries, longer for evergreen topics.
- Limit Results to avoid overwhelming the Agent with too many irrelevant links.
👉 With the Web Search Tool, your Agent extends beyond static knowledge and can operate with real-time awareness, enhancing its usefulness in dynamic scenarios like support, sales, and research.