Skip to main content

Researcher Agent

The Researcher Agent conducts web research on specified topics and provides comprehensive summaries of findings, supporting the writing process with accurate information.

Overview​

This agent performs targeted research to gather information, verify facts, and provide context for your book's content. It combines AI-generated summaries with web scraping capabilities to deliver comprehensive research results.

Features​

Research Capabilities​

  • Web content analysis
  • Search result processing
  • Content summarization
  • Source verification
  • Data compilation

Data Collection​

  • Web scraping
  • Search optimization
  • Content filtering
  • Source ranking
  • Result aggregation

Output Generation​

  • Detailed summaries
  • Source citations
  • Key findings
  • Relevant quotes
  • Research context

Research Process​

  1. Query Processing

    • Analyze research needs
    • Optimize search terms
    • Define scope
    • Identify key areas

  2. Data Collection

    • Web scraping
    • Content gathering
    • Source verification
    • Information filtering

  3. Analysis & Synthesis

    • Content summarization
    • Data organization
    • Finding compilation
    • Report generation

Output Format​

Research results are provided in Markdown format:

# Research Report: [Topic]

## AI-Generated Summary
[Comprehensive overview of findings]

## Web Search Results
### [Source Title]
- URL: [Source URL]
- Key Points:
  - [Relevant information]
  - [Supporting details]

### [Additional Sources...]

Usage​

# Through Project Manager
libriscribe research "research query"

# Direct usage
researcher.execute(query, output_path)

Best Practices​

  1. Query Optimization

    • Use specific terms
    • Define scope clearly
    • Include context
    • Specify requirements

  2. Source Management

    • Verify reliability
    • Track citations
    • Document sources
    • Rate credibility

  3. Content Processing

    • Filter relevance
    • Organize findings
    • Summarize effectively
    • Maintain context

Integration​

Works with:

  • Fact Checker for verification
  • Content Reviewer for context
  • Editor Agent for integration
  • Project Manager for coordination

Error Handling​

Manages:

  • Failed searches
  • Invalid sources
  • Processing errors
  • Connection issues

Configuration​

Customizable for:

  • Search depth
  • Source requirements
  • Result filtering
  • Output format

Security & Ethics​

  • Respects robots.txt
  • Implements rate limiting
  • Follows scraping ethics
  • Maintains attribution

Downstream Usage in Chapter Writing​

The research_results.md file produced by this agent is automatically read by the Chapter Writer Agent when generating each scene. The first 1 200 characters are injected as a RESEARCH CONTEXT block in the scene prompt, providing the LLM with factual grounding from your research.

Run the Researcher before or during chapter writing for best results:

libriscribe research "your research topic here"

The research file persists across sessions — re-run to update it with new findings. See Chapter Writer Agent for details on how it is used.