Web Research

Structured web research patterns using search APIs, content extraction, and synthesis. For agents that need to gather and analyze online information.

grandamenium c052442 1.2 KB Updated

File contents

Web Research

Patterns for agents performing structured web research.

Search Strategy

  1. Break the research question into sub-queries
  2. Search each sub-query independently
  3. Extract relevant content from top results
  4. Synthesize findings with source attribution

Search Tools

Use WebSearch for broad queries, WebFetch for specific URLs:

WebSearch: "cortextOS multi-agent orchestration 2026"
WebFetch: "https://docs.anthropic.com/en/docs/agents"

Content Extraction

  • Extract key facts, quotes, and data points
  • Note the source URL for each finding
  • Check publication dates for freshness
  • Cross-reference claims across multiple sources

Research Output Format

## Finding: [Topic]
**Source:** [URL]
**Date:** [Publication date]
**Key points:**
- Point 1
- Point 2

**Relevance:** [How this relates to the research question]

Best Practices

  • Always attribute sources
  • Prefer primary sources over secondary
  • Check for recency (information may be outdated)
  • Synthesize, don't just aggregate
  • Flag conflicting information explicitly

grandamenium/cortextos/tree/main/skills/web-research commit c052442b6b

Frequently asked questions

npx skillmds@latest add grandamenium/web-research