Back to Details

deep-research

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Deep Research

深度研究

This skill conducts comprehensive research on complex topics using a multi-agent architecture, producing detailed reports similar to academic journals or whitepapers.
本Skill采用多Agent架构对复杂主题开展全面研究，生成类似学术期刊或白皮书的详细报告。

Purpose

用途

The deep-research skill transforms broad research questions into thorough, well-cited reports using a three-agent system:
  1. Lead Agent (You): Conducts interviews, plans research, orchestrates subagents
  2. Researcher Agents: Execute web searches and save findings to files
  3. Report-Writer Agent: Synthesizes research notes into final report
深度研究Skill通过三Agent系统将宽泛的研究问题转化为详尽且引用规范的报告：
  1. 主导Agent（你）：开展需求访谈、规划研究、协调子Agent
  2. 研究员Agent：执行网络搜索并将研究结果保存至文件
  3. 报告撰写Agent：整合研究笔记生成最终报告

When to Use This Skill

使用场景

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
在用户提出以下需求时使用本Skill：
  • 对复杂主题进行深度研究
  • 需要一份全面的报告或分析内容
  • 研究需整合多来源信息并进行综合分析
  • 达到学术或白皮书标准的深度调研
  • 带有规范引用的详细分析
请勿在以下场景使用本Skill：
  • 简单的事实查询
  • 单一来源的信息检索
  • 仅针对代码仓库的研究
  • 快速的探索性搜索

Agent Architecture

Agent架构

Lead Agent (You - the Orchestrator)

主导Agent（你 - 协调者）

Role: Interview user, plan research threads, spawn and coordinate subagents
Tools allowed: Task (to spawn subagents), AskUserQuestion, Write (for research plan only)
Responsibilities:
  • Conduct user interview to scope research
  • Perform initial reconnaissance
  • Decompose topic into 10+ research threads
  • Spawn researcher agents in parallel
  • Spawn report-writer agent after research completes
角色：与用户访谈、规划研究方向、创建并协调子Agent
允许使用的工具：Task（用于创建子Agent）、AskUserQuestion、Write（仅用于撰写研究计划）
职责：
  • 与用户访谈以明确研究范围
  • 执行初步调研
  • 将研究主题拆解为10个以上的研究方向
  • 并行创建多个研究员Agent
  • 研究完成后创建报告撰写Agent

Researcher Agents

研究员Agent

Role: Execute focused research on assigned subtopic
Tools allowed: WebSearch, WebFetch, Write
Responsibilities:
  • Search the web for information on assigned topic
  • Fetch and analyze relevant pages
  • Save structured research notes to 
    research_notes/
    directory
Output format: Each researcher saves a markdown file to 
research_notes/[subtopic-slug].md
with:
  • Summary of findings
  • Key facts and data points
  • Source URLs with brief descriptions
  • Notable quotes or excerpts
  • Conflicts or gaps identified
角色：针对指定子主题开展针对性研究
允许使用的工具：WebSearch、WebFetch、Write
职责：
  • 针对指定主题进行网络搜索
  • 获取并分析相关页面内容
  • 将结构化的研究笔记保存至
    research_notes/
    目录
输出格式：每个研究员需将Markdown文件保存至
research_notes/[subtopic-slug].md
，文件包含：
  • 研究结果摘要
  • 关键事实和数据点
  • 带简要说明的来源URL
  • 重要引用或摘录
  • 发现的矛盾点或研究空白

Report-Writer Agent

报告撰写Agent

Role: Synthesize all research notes into final report
Tools allowed: Read, Glob, Write
Responsibilities:
  • Read all files from 
    research_notes/
    directory
  • Identify themes, patterns, and conflicts across sources
  • Structure and write the final report
  • Create the sources bibliography
角色：整合所有研究笔记生成最终报告
允许使用的工具：Read、Glob、Write
职责：
  • 读取
    research_notes/
    目录下的所有文件
  • 识别跨来源的主题、模式和矛盾点
  • 构建并撰写最终报告
  • 创建参考文献目录

Research Process

研究流程

Phase 1: Interview and Scope Definition

阶段1：访谈与范围定义

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.
首先与用户访谈以了解其研究需求，需询问以下问题：
  1. 研究目标：用户希望了解或决定什么？
  2. 深度与广度：研究需达到何种全面程度？
  3. 目标受众：谁会阅读这份报告？
  4. 核心问题：需要解答哪些具体问题？
  5. 时间限制：研究内容是否有时间敏感性？
  6. 范围边界：明确需要包含或排除的内容
访谈需全面且高效，使用AskUserQuestion工具，最多通过2-3轮提问收集信息。

Phase 2: Initial Reconnaissance

阶段2：初步调研

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
Save the research plan to 
research_plan.md
documenting:
  • The research threads identified
  • Which researcher will handle each thread
  • Expected output from each researcher
访谈完成后，执行初步调研以明确研究格局：
  1. 进行3-5次宽泛的网络搜索以梳理主题领域
  2. 识别关键子主题、领域和重点方向
  3. 记录有价值的来源、权威观点和研究空白
  4. 制定包含10个以上具体研究方向的研究计划
将研究计划保存至
research_plan.md
，需记录：
  • 确定的研究方向
  • 每个研究方向对应的研究员Agent
  • 每个研究员的预期输出

Phase 3: Parallel Research (Researcher Agents)

阶段3：并行研究（研究员Agent）

Launch 10+ researcher agents in parallel using the Task tool. Each agent receives a focused research assignment.
Spawning researcher agents:
Task tool with:
- subagent_type: "general-purpose"
- prompt: Include these elements:
  1. Clear statement: "You are a RESEARCHER agent"
  2. Specific subtopic assignment
  3. Tool restrictions: "Only use WebSearch, WebFetch, and Write tools"
  4. Output instructions: "Save your findings to research_notes/[subtopic].md"
  5. Format requirements for the research notes file
Example researcher prompt:
You are a RESEARCHER agent investigating: "Technical implementation of quantum error correction"

YOUR TOOLS: Only use WebSearch, WebFetch, and Write.

TASK:
1. Use WebSearch to find authoritative sources on quantum error correction implementation
2. Use WebFetch to extract detailed information from promising sources
3. Save your findings to research_notes/quantum-error-correction.md

OUTPUT FORMAT (save to research_notes/quantum-error-correction.md):
使用Task工具并行创建10个以上的研究员Agent，每个Agent分配针对性的研究任务。
创建研究员Agent的方式：
Task工具参数：
- subagent_type: "general-purpose"
- prompt: 包含以下元素：
  1. 明确声明："You are a RESEARCHER agent"
  2. 具体子主题分配
  3. 工具限制："Only use WebSearch, WebFetch, and Write tools"
  4. 输出说明："Save your findings to research_notes/[subtopic].md"
  5. 研究笔记文件的格式要求
研究员提示词示例：
You are a RESEARCHER agent investigating: "Technical implementation of quantum error correction"

YOUR TOOLS: Only use WebSearch, WebFetch, and Write.

TASK:
1. Use WebSearch to find authoritative sources on quantum error correction implementation
2. Use WebFetch to extract detailed information from promising sources
3. Save your findings to research_notes/quantum-error-correction.md

OUTPUT FORMAT (save to research_notes/quantum-error-correction.md):

Quantum Error Correction Implementation

Quantum Error Correction Implementation

Summary

Summary

[2-3 paragraph summary of key findings]
[2-3 paragraph summary of key findings]

Key Findings

Key Findings

  • [Bullet points of important facts, data, techniques]
  • [Bullet points of important facts, data, techniques]

Sources

Sources

  1. [URL] - [Brief description of what this source contributed]
  2. [URL] - [Brief description] ...
  1. [URL] - [Brief description of what this source contributed]
  2. [URL] - [Brief description] ...

Notable Quotes

Notable Quotes

"[Relevant quote]" - Source
"[Relevant quote]" - Source

Gaps and Conflicts

Gaps and Conflicts

  • [Any conflicting information or areas needing more research]

**Launch all researcher agents in a single message** with multiple Task tool calls for true parallelism.
  • [Any conflicting information or areas needing more research]

**在单条消息中发起所有研究员Agent的创建请求**，通过多个Task工具调用实现真正的并行处理。

Phase 4: Report Generation (Report-Writer Agent)

阶段4：报告生成（报告撰写Agent）

After all researcher agents complete, spawn a single report-writer agent:
Spawning the report-writer agent:
Task tool with:
- subagent_type: "general-purpose"
- prompt: Include these elements:
  1. Clear statement: "You are a REPORT-WRITER agent"
  2. Tool restrictions: "Only use Read, Glob, and Write tools"
  3. Instructions to read all files from research_notes/
  4. Report structure requirements
  5. Output file paths for report and sources
Example report-writer prompt:
You are a REPORT-WRITER agent synthesizing research findings into a final report.

YOUR TOOLS: Only use Read, Glob, and Write.

TASK:
1. Use Glob to list all files in research_notes/
2. Use Read to load each research notes file
3. Synthesize findings into a comprehensive report
4. Write the final report to [topic]-report.md
5. Write the sources bibliography to [topic]-sources.md

REPORT STRUCTURE:
- Executive Summary (2-3 paragraphs)
- [Adaptive middle sections based on topic]
- Critical Analysis
- Conclusions
- References (numbered citations)

SOURCES FILE STRUCTURE:
所有研究员Agent完成任务后，创建单个报告撰写Agent：
创建报告撰写Agent的方式：
Task工具参数：
- subagent_type: "general-purpose"
- prompt: 包含以下元素：
  1. 明确声明："You are a REPORT-WRITER agent"
  2. 工具限制："Only use Read, Glob, and Write tools"
  3. 读取research_notes/目录下所有文件的指令
  4. 报告结构要求
  5. 报告和参考文献的输出文件路径
报告撰写Agent提示词示例：
You are a REPORT-WRITER agent synthesizing research findings into a final report.

YOUR TOOLS: Only use Read, Glob, and Write.

TASK:
1. Use Glob to list all files in research_notes/
2. Use Read to load each research notes file
3. Synthesize findings into a comprehensive report
4. Write the final report to [topic]-report.md
5. Write the sources bibliography to [topic]-sources.md

REPORT STRUCTURE:
- Executive Summary (2-3 paragraphs)
- [Adaptive middle sections based on topic]
- Critical Analysis
- Conclusions
- References (numbered citations)

SOURCES FILE STRUCTURE:

Research Sources for [Topic]

Research Sources for [Topic]

[1] Source Title

[1] Source Title

  • URL: [url]
  • Accessed: [date]
  • Type: [Academic paper / Blog post / Documentation / News article]
  • Key Points: [bullet points]
  • Relevance: [why this source matters]
WRITING GUIDELINES:
  • Use numbered citations [1], [2], etc.
  • Cross-reference findings across multiple researcher notes
  • Note any conflicts or gaps in the research
  • Use clear, precise academic language
  • Include tables for comparisons where appropriate
undefined
  • URL: [url]
  • Accessed: [date]
  • Type: [Academic paper / Blog post / Documentation / News article]
  • Key Points: [bullet points]
  • Relevance: [why this source matters]
WRITING GUIDELINES:
  • Use numbered citations [1], [2], etc.
  • Cross-reference findings across multiple researcher notes
  • Note any conflicts or gaps in the research
  • Use clear, precise academic language
  • Include tables for comparisons where appropriate
undefined

Phase 5: Output and Summary

阶段5：输出与总结

After the report-writer completes:
  1. Inform the user of the generated files:
    • [topic]-report.md
      : Main research report
    • [topic]-sources.md
      : Complete bibliography
    • research_notes/
      : Directory of raw research (can be deleted)
  2. Provide a brief verbal summary of key findings
  3. Offer to answer follow-up questions or expand on any section
报告撰写Agent完成任务后：
  1. 告知用户生成的文件：
    • [topic]-report.md
      ：主研究报告
    • [topic]-sources.md
      ：完整参考文献
    • research_notes/
      ：原始研究笔记目录（可删除）
  2. 口头简要总结关键研究结果
  3. 主动提出可解答后续问题或扩展任意章节

File Structure

文件结构

./
├── research_plan.md           # Your research plan (Phase 2)
├── research_notes/            # Researcher agent outputs (Phase 3)
│   ├── subtopic-1.md
│   ├── subtopic-2.md
│   └── ...
├── [topic]-report.md          # Final report (Phase 4)
└── [topic]-sources.md         # Bibliography (Phase 4)
./
├── research_plan.md           # 你的研究计划（阶段2）
├── research_notes/            # 研究员Agent输出（阶段3）
│   ├── subtopic-1.md
│   ├── subtopic-2.md
│   └── ...
├── [topic]-report.md          # 最终报告（阶段4）
└── [topic]-sources.md         # 参考文献（阶段4）

Logging and Observability

日志与可观测性

Track research progress by documenting in 
research_plan.md
:
  1. Research threads assigned: List each subtopic and its researcher
  2. Status tracking: Note when each researcher completes
  3. Issues encountered: Document any gaps or conflicts found
This provides transparency into the research process and helps with debugging or expanding research later.
通过在
research_plan.md
中记录内容跟踪研究进度：
  1. 已分配的研究方向：列出每个子主题及其对应的研究员
  2. 状态跟踪：记录每个研究员的完成时间
  3. 遇到的问题：记录发现的研究空白或矛盾点
这能让研究过程透明化，便于后续调试或扩展研究。

Best Practices

最佳实践

Agent Separation

Agent职责分离

  • Lead agent: ONLY spawns agents and coordinates - no direct research
  • Researchers: ONLY search, fetch, and write notes - no synthesis
  • Report-writer: ONLY reads notes and writes report - no new research
This separation ensures clean handoffs and reproducible results.
  • 主导Agent：仅负责创建和协调Agent，不直接开展研究
  • 研究员Agent：仅负责搜索、获取和撰写笔记，不进行内容整合
  • 报告撰写Agent：仅负责读取笔记和撰写报告，不开展新研究
这种职责分离确保了清晰的工作交接和可复现的结果。

Research Quality

研究质量

  • Prioritize authoritative, recent sources (especially for time-sensitive topics)
  • Cross-reference claims across multiple researcher notes
  • Note conflicting information or perspectives
  • Distinguish between facts, expert opinions, and speculation
  • Be transparent about limitations in available information
  • 优先选择权威、近期的来源（尤其是对时间敏感的主题）
  • 跨多个研究员笔记交叉验证主张
  • 记录矛盾的信息或观点
  • 区分事实、专家意见和推测
  • 对可用信息的局限性保持透明

Efficiency

效率

  • Launch all researcher agents truly in parallel (single message, multiple Task tool calls)
  • Use model="haiku" for researcher agents to reduce costs
  • Use model="sonnet" for report-writer agent for better synthesis
  • Clear task delineation prevents redundant research
  • 真正并行启动所有研究员Agent（单条消息，多个Task工具调用）
  • 为研究员Agent使用model="haiku"以降低成本
  • 为报告撰写Agent使用model="sonnet"以提升整合效果
  • 明确的任务划分避免重复研究

Common Patterns

常见模式

Comparative Research

对比研究

When comparing technologies, approaches, or solutions:
  • Assign one researcher per option being compared
  • Assign one researcher for cross-cutting concerns (performance, cost, etc.)
  • Report-writer creates comparison tables
当对比技术、方法或解决方案时：
  • 为每个待对比选项分配一名研究员
  • 为跨领域关注点（性能、成本等）分配一名研究员
  • 报告撰写Agent创建对比表格

Technical Deep-Dives

技术深度调研

When researching technical topics:
  • Assign researchers to: fundamentals, implementation, case studies, limitations
  • Report-writer structures from basics to advanced
当研究技术主题时：
  • 为以下方向分配研究员：基础原理、实现方式、案例研究、局限性
  • 报告撰写Agent按照从基础到进阶的结构组织内容

Market/Landscape Research

市场/格局研究

When surveying a domain or market:
  • Assign researchers to: major players, emerging players, trends, analysis firms
  • Report-writer categorizes and evaluates the landscape
当调研某个领域或市场时：
  • 为以下方向分配研究员：主要参与者、新兴参与者、趋势、分析机构
  • 报告撰写Agent对格局进行分类和评估

Historical/Evolution Research

历史/演进研究

When investigating how something developed:
  • Assign researchers to different time periods or key events
  • Report-writer creates timeline and connects to present
当研究事物的发展历程时：
  • 为不同时间段或关键事件分配研究员
  • 报告撰写Agent创建时间线并关联当前现状