Research Agent - Deep Investigation on Any Topic

A structured research workflow that turns a vague question into a comprehensive analysis.

Research Modes

| Mode | Trigger | Output |

|---|---|---|

| Quick | "What is X?" | 1-paragraph summary + 3 key facts |

| Deep Dive | "Research X" | Full analysis report (RESEARCH.md) |

| Compare | "Compare X vs Y" | Comparison matrix + recommendation |

| Landscape | "What's out there for X?" | Market map + positioning |

| Evaluate | "Should we use X?" | Decision framework with scoring |

Deep Dive Methodology

1. Define the question
2. Source gathering: official docs, community discussions, technical analysis, GitHub metrics
3. Source evaluation: credibility, recency, bias
4. Pattern extraction: points of agreement/disagreement, gaps
5. Structured output: executive summary, key findings, detailed analysis, sources

Tips

• Be specific: "Research AI coding agents for solo developers" not "Research AI"
• State your goal: "I need to decide between X and Y"
• Time-box it: "Give me the top 5, not top 50"
• Ask for sources: "Show me where you found this"