Web Summary

Description

Searches the web for any topic using the DuckDuckGo Instant Answer API, fetches up to maxSources related pages, and returns a structured summary with extracted key facts and source URLs.

Input Schema

{
  "type": "object",
  "properties": {
    "topic": {
      "type": "string",
      "description": "The topic or question to research"
    },
    "maxSources": {
      "type": "number",
      "description": "Maximum number of web sources to fetch (1–10, default 3)"
    }
  },
  "required": ["topic"]
}

Output Schema

{
  "type": "object",
  "properties": {
    "summary":   { "type": "string",  "description": "Paragraph summary of the topic" },
    "sources":   { "type": "array",   "items": { "type": "string" }, "description": "URLs consulted" },
    "keyFacts":  { "type": "array",   "items": { "type": "string" }, "description": "Up to 8 extracted facts" }
  }
}

Example

const result = await registry.execute('research.web-summary', {
  topic: 'quantum computing breakthroughs 2025',
  maxSources: 5,
});
// result.data.summary   → paragraph text
// result.data.sources   → ['https://...', ...]
// result.data.keyFacts  → ['Fact 1.', 'Fact 2.', ...]

Notes

• Uses DuckDuckGo Instant Answer (no API key required).
• Falls back gracefully if pages are unreachable.
• No LLM calls — pure web scraping and text extraction.
• Timeout: 60 seconds total.