Deep Research

This skill conducts comprehensive research on complex topics, producing detailed reports similar to academic journals or whitepapers.

Purpose

The deep-research skill transforms broad research questions into thorough, well-cited reports by:

1. Conducting structured interviews to understand research goals
2. Performing iterative deepening to identify key areas
3. Launching parallel research subagents for comprehensive coverage
4. Synthesizing findings into a cohesive, academically-styled report
5. Maintaining a separate bibliography with all sources

When to Use This Skill

Use this skill when the user requests:

• In-depth research on a complex topic
• A comprehensive report or analysis
• Research that requires multiple sources and synthesis
• Deep investigation similar to academic or whitepaper standards
• Detailed analysis with proper citations

Do NOT use this skill for:

• Simple fact-finding queries
• Single-source information lookup
• Code-only research within repositories
• Quick exploratory searches

Research Process

Phase 1: Interview and Scope Definition

Start by interviewing the user to understand their research needs. Ask questions about:

1. Research objectives: What are they trying to understand or decide?
2. Depth and breadth: How comprehensive should the research be?
3. Target audience: Who will read this report?
4. Key questions: What specific questions need answering?
5. Time constraints: Is this time-sensitive information?
6. Scope boundaries: What should be explicitly included or excluded?

The interview should be thorough but efficient. Use the AskUserQuestion tool to gather this information in 2-3 rounds of questions maximum.

Phase 2: Initial Reconnaissance (Iterative Deepening)

After the interview, perform initial reconnaissance to identify the research landscape:

1. Conduct 3-5 broad web searches to map the topic space
2. Identify key subtopics, domains, and areas of focus
3. Note promising sources, authoritative voices, and research gaps
4. Create a research plan outlining 10+ specific research threads

This phase should be relatively quick (5-10 searches) but strategic. The goal is to create an informed plan for the parallel exploration phase.

Document findings in working notes but do not create the final report yet.

Phase 3: Parallel Exploration (High Parallelism)

Launch 10+ parallel research subagents using the Task tool with subagent_type="general-purpose". Each agent should investigate a specific research thread identified in Phase 2.

Agent assignment strategy:

• Assign each agent a focused research question or subtopic
• Ensure agents have clear, non-overlapping objectives
• Instruct agents to use WebSearch and WebFetch tools extensively
• Request that each agent return structured findings with sources

Example agent tasks:

• "Research the technical implementation of [specific approach]"
• "Investigate the historical context and evolution of [topic]"
• "Compare and contrast [approach A] vs [approach B]"
• "Analyze the current state of research on [subtopic]"
• "Identify key challenges and limitations in [area]"

Launch all agents in parallel (in a single message with multiple Task tool calls) for maximum efficiency.

Phase 4: Synthesis and Report Generation

After all subagents complete:

1. Review all findings from the parallel research phase
2. Identify common themes, conflicts, and key insights
3. Structure the report using a hybrid format (see below)
4. Write the report with academic rigor and proper citations
5. Create a separate sources bibliography file

Report structure (hybrid format):

The report should always include these core sections:

• Executive Summary: 2-3 paragraph overview of key findings
• [Adaptive Middle Sections]: Structure based on topic (comparisons, historical analysis, technical deep-dives, etc.)
• Critical Analysis: Deep evaluation, synthesis, and interpretation
• Conclusions: Summary of findings and implications
• References: Numbered citations used throughout

The middle sections should adapt to the research topic:

• For comparative research: Side-by-side analysis sections
• For technical topics: Architecture, implementation, tradeoffs sections
• For historical topics: Timeline, evolution, impact sections
• For survey research: Landscape, categories, evaluation sections

Citation style:

• Use numbered citations in the text: [1], [2], etc.
• Include inline source context when relevant: "According to Smith et al. [3], ..."
• Maintain a complete references section at the end
• Create a separate sources-bibliography.md file with full source details

Phase 5: Output

Save two files in the current working directory:

1. [topic-name]-report.md: The main research report
2. [topic-name]-sources.md: Complete bibliography with:
   • Full URLs
   • Access dates
   • Source descriptions
   • Key excerpts or quotes
   • Relevance notes

Use clear, descriptive filenames based on the research topic (e.g., "quantum-computing-hardware-report.md").

Inform the user of the file locations and provide a brief summary of the research findings.

Best Practices

Research Quality

• Prioritize authoritative, recent sources (especially for time-sensitive topics)
• Cross-reference claims across multiple sources
• Note conflicting information or perspectives
• Distinguish between facts, expert opinions, and speculation
• Be transparent about limitations in available information

Writing Style

• Use clear, precise academic language
• Define technical terms and acronyms on first use
• Provide context and background for complex concepts
• Use structured formatting (headers, lists, tables) for readability
• Include data, statistics, and concrete examples where relevant

Source Management

• Maintain meticulous source tracking throughout research
• Cite sources immediately when incorporating information
• Prefer primary sources over secondary when possible
• Include diverse perspectives and sources
• Verify critical claims across multiple sources

Efficiency

• Launch agents truly in parallel (single message, multiple tool calls)
• Use model="haiku" for subagents when appropriate for cost savings
• Avoid redundant research between agents through clear task delineation
• Work iteratively: reconnaissance → parallel research → synthesis

Common Patterns

Comparative Research

When comparing technologies, approaches, or solutions:

1. Research each option thoroughly in parallel
2. Create structured comparison sections (features, performance, costs, tradeoffs)
3. Use tables for side-by-side comparisons
4. Provide clear recommendations or trade-off analysis

Technical Deep-Dives

When researching technical topics:

1. Start with fundamentals and key concepts
2. Progress to implementation details and architecture
3. Cover real-world applications and case studies
4. Address limitations, challenges, and future directions

Market/Landscape Research

When surveying a domain or market:

1. Categorize the landscape (player types, segments, approaches)
2. Profile key players or solutions
3. Identify trends and patterns
4. Analyze implications and future outlook

Historical/Evolution Research

When investigating how something developed:

1. Establish timeline and key milestones
2. Identify driving forces and catalysts
3. Analyze impact and consequences
4. Connect historical context to current state

References

Store detailed source information in the references file:

# Research Sources for [Topic]

## [1] Source Title
- **URL**: https://example.com/article
- **Accessed**: 2025-11-13
- **Type**: Academic paper / Blog post / Documentation / News article
- **Key Points**:
  - Main finding or claim 1
  - Main finding or claim 2
- **Relevance**: Why this source matters to the research

## [2] Source Title
...

This allows the main report to remain clean while preserving full source details for verification and further research.