Gemini Researcher

Deep research capability using Google's Gemini CLI for comprehensive analysis.

When to Use

• •Investigating technical topics, APIs, or frameworks
• •Validating architectural decisions with evidence
• •Gathering market intelligence or competitive analysis
• •Researching compliance, security, or best practices

Prerequisites

• •Gemini CLI installed: pip install google-generativeai or via Google Cloud
• •API key configured: export GOOGLE_API_KEY='your-key'

Workflow

Quick Research

gemini -p "Research [topic]: goals, current state, recommendations" > research-notes.md

Structured Research

1. •

   Craft a focused prompt with clear deliverables:

   code

   Research [TOPIC]:
   - Current state of the art
   - Key players and solutions
   - Pros/cons analysis
   - Actionable recommendations
   Include sources with URLs.
   

2. •

   Execute: gemini -p "your prompt" > docs/research/YYYY-MM-topic.md

3. •

   Review and structure output:

   • •Summary (2-3 sentences)
   • •Key Findings (bullet points)
   • •Sources (with URLs and dates)
   • •Next Actions (specific tasks)
4. •

   Link findings to relevant project documentation

Prompt Templates

Technology Evaluation

Evaluate [TECHNOLOGY] for [USE CASE]:
1. What problems does it solve?
2. Alternatives and comparisons
3. Integration complexity
4. Community and support status
5. Recommendation with reasoning

Best Practices Research

Research best practices for [DOMAIN]:
- Industry standards
- Common pitfalls
- Implementation patterns
- Tools and frameworks
Include recent sources (2024-2025).

Tips

• •Always include sources with URLs; note confidence when Gemini hedges
• •For lengthy output, distill into bullet-friendly sections
• •Cross-reference with multiple queries for controversial topics
• •Save raw output before editing for reference