research-workflow
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseResearch Workflow
研究工作流
A structured methodology for conducting comprehensive research. This skill guides you through planning, executing, analyzing, and synthesizing research on any topic.
一种用于开展全面研究的结构化方法。该Skill将引导你完成任意主题的研究规划、执行、分析和信息整合全流程。
When to Use This Skill
何时使用该Skill
Use this skill when:
- The user needs comprehensive research on a topic
- Multiple search queries are needed to fully answer a question
- Source credibility and synthesis matter
- A research report or documented findings are expected
- Keywords mentioned: research, investigate, deep dive, comprehensive analysis
Do NOT use this skill when:
- A single quick search will suffice (use web-search instead)
- The user just wants a simple fact lookup
- No synthesis or analysis is needed
- Time is extremely limited
在以下场景使用该Skill:
- 用户需要针对某个主题进行全面研究
- 需要多个搜索查询才能完整解答问题
- 来源可信度和信息整合至关重要
- 需要产出研究报告或记录研究发现
- 提及以下关键词:研究、调查、深度探究、全面分析
请勿在以下场景使用该Skill:
- 单次快速搜索即可满足需求(请使用web-search替代)
- 用户仅需查找简单事实
- 无需进行信息整合或分析
- 时间极为有限
Prerequisites
前置条件
Before using this skill, ensure:
- Web search capability is available (web-search skill, WebSearch tool, or similar)
- Sufficient time for multi-phase research process
- Clear understanding of the research question or topic
使用该Skill前,请确保:
- 具备网页搜索能力(拥有web-search Skill、WebSearch工具或类似功能)
- 有充足时间完成多阶段研究流程
- 清晰理解研究问题或主题
Research Phases Overview
研究阶段概述
┌─────────────────────────────────────────────────────────────┐
│ RESEARCH WORKFLOW │
├─────────────────────────────────────────────────────────────┤
│ │
│ 1. PLANNING 2. EXECUTION │
│ ┌──────────────┐ ┌──────────────┐ │
│ │ Define │ │ Run searches │ │
│ │ questions │───>│ Evaluate │ │
│ │ Plan queries │ │ sources │ │
│ └──────────────┘ └──────────────┘ │
│ │ │ │
│ v v │
│ 3. ANALYSIS 4. SYNTHESIS │
│ ┌──────────────┐ ┌──────────────┐ │
│ │ Organize │ │ Create │ │
│ │ findings │───>│ coherent │ │
│ │ Find patterns│ │ output │ │
│ └──────────────┘ └──────────────┘ │
│ │
└─────────────────────────────────────────────────────────────┘┌─────────────────────────────────────────────────────────────┐
│ RESEARCH WORKFLOW │
├─────────────────────────────────────────────────────────────┤
│ │
│ 1. PLANNING 2. EXECUTION │
│ ┌──────────────┐ ┌──────────────┐ │
│ │ Define │ │ Run searches │ │
│ │ questions │───>│ Evaluate │ │
│ │ Plan queries │ │ sources │ │
│ └──────────────┘ └──────────────┘ │
│ │ │ │
│ v v │
│ 3. ANALYSIS 4. SYNTHESIS │
│ ┌──────────────┐ ┌──────────────┐ │
│ │ Organize │ │ Create │ │
│ │ findings │───>│ coherent │ │
│ │ Find patterns│ │ output │ │
│ └──────────────┘ └──────────────┘ │
│ │
└─────────────────────────────────────────────────────────────┘Phase 1: Planning
阶段1:规划
Before any searches, establish a clear research plan.
在进行任何搜索前,先制定清晰的研究计划。
Step 1: Define the Research Question
步骤1:定义研究问题
Convert the topic into specific, answerable questions.
Example:
- Topic: "AI in healthcare"
- Questions:
- What are the current applications of AI in healthcare?
- What are the main benefits and challenges?
- What regulations govern AI in healthcare?
- What are the latest developments (last 6 months)?
将主题转化为具体、可解答的问题。
示例:
- 主题:"医疗领域中的AI"
- 问题:
- AI在医疗领域的当前应用有哪些?
- 主要优势和挑战是什么?
- 哪些法规监管医疗领域的AI?
- 最新发展(过去6个月)有哪些?
Step 2: Identify Sub-Topics
步骤2:确定子主题
Break down the main topic into searchable components:
- Core concepts and definitions
- Current state and applications
- Benefits and advantages
- Challenges and limitations
- Recent developments
- Future trends
将主主题拆解为可搜索的组成部分:
- 核心概念与定义
- 当前现状与应用
- 优势与益处
- 挑战与局限
- 最新发展
- 未来趋势
Step 3: Plan Search Strategy
步骤3:规划搜索策略
Create a search plan with query progression:
-
Broad queries first: Get overall landscape
- "[topic] overview"
- "[topic] introduction guide"
-
Specific queries next: Dive into details
- "[topic] specific aspect"
- "[topic] case study"
-
Verification queries last: Confirm findings
- "[topic] criticism challenges"
- "[topic] latest news [year]"
Use the template at assets/research-plan-template.md to document your plan.
制定包含查询递进的搜索计划:
-
先进行宽泛查询:了解整体概况
- "[主题] 概述"
- "[主题] 入门指南"
-
再进行具体查询:深入细节
- "[主题] 特定方面"
- "[主题] 案例研究"
-
最后进行验证查询:确认研究发现
- "[主题] 批评与挑战"
- "[主题] [年份] 最新资讯"
使用[assets/research-plan-template.md]中的模板记录你的计划。
Phase 2: Execution
阶段2:执行
Execute your search plan systematically.
系统地执行你的搜索计划。
Step 1: Run Searches
步骤1:执行搜索
Execute queries in order, using appropriate search parameters:
bash
undefined按顺序执行查询,使用合适的搜索参数:
bash
undefinedBroad overview
宽泛概述
web-search "AI in healthcare overview 2024"
web-search "AI in healthcare overview 2024"
Specific deep dive
深度探究细节
web-search "AI diagnostic imaging applications" --depth advanced
web-search "AI diagnostic imaging applications" --depth advanced
Current news
当前资讯
web-search "AI healthcare regulations 2024" --topic news --time month
undefinedweb-search "AI healthcare regulations 2024" --topic news --time month
undefinedStep 2: Document Each Search
步骤2:记录每次搜索
For each search, record:
- Query used
- Number of results reviewed
- Key findings (2-3 bullet points)
- Notable sources
- New questions raised
针对每次搜索,记录:
- 使用的查询语句
- 查看的结果数量
- 关键发现(2-3个要点)
- 值得关注的来源
- 新产生的问题
Step 3: Evaluate Sources
步骤3:评估来源
Use the checklist at assets/source-evaluation-checklist.md to assess:
Credibility Indicators:
- Author/organization expertise
- Publication reputation
- Date of publication
- Citations and references
Quality Signals:
- Evidence-based claims
- Multiple perspectives
- Clear methodology
使用[assets/source-evaluation-checklist.md]中的清单评估:
可信度指标:
- 作者/机构的专业程度
- 出版物的声誉
- 发布日期
- 引用与参考文献
质量信号:
- 基于证据的主张
- 多视角呈现
- 清晰的研究方法
Step 4: Iterate as Needed
步骤4:按需迭代
Research is not linear. Based on findings:
- Add new queries for gaps discovered
- Verify surprising claims
- Explore unexpected connections
研究并非线性流程。根据研究发现:
- 针对发现的信息缺口添加新查询
- 验证意外的研究结果
- 探索未预期的关联
Phase 3: Analysis
阶段3:分析
Organize and analyze your collected findings.
整理并分析收集到的研究发现。
Step 1: Group Findings by Theme
步骤1:按主题分组研究发现
Organize results into categories:
- Core concepts
- Current state
- Benefits/opportunities
- Challenges/risks
- Recent developments
- Expert opinions
将结果分类整理:
- 核心概念
- 当前现状
- 优势/机遇
- 挑战/风险
- 最新发展
- 专家观点
Step 2: Identify Patterns
步骤2:识别模式
Look for:
- Consensus: Where do multiple sources agree?
- Conflicts: Where do sources disagree?
- Gaps: What questions remain unanswered?
- Trends: What direction is the field moving?
寻找以下内容:
- 共识:多个来源达成一致的点?
- 冲突:来源之间存在分歧的点?
- 缺口:哪些问题仍未得到解答?
- 趋势:该领域的发展方向是什么?
Step 3: Assess Confidence
步骤3:评估可信度
For each finding, determine confidence level:
- High: Multiple authoritative sources agree
- Medium: Some evidence, limited sources
- Low: Single source or conflicting information
针对每个研究发现,确定可信度等级:
- 高:多个权威来源达成一致
- 中:有部分证据,但来源有限
- 低:单一来源或信息存在冲突
Step 4: Note Limitations
步骤4:记录局限性
Document:
- What couldn't be found
- Areas needing more research
- Potential biases in sources
记录:
- 无法找到的信息
- 需要进一步研究的领域
- 来源中可能存在的偏见
Phase 4: Synthesis
阶段4:信息整合
Create coherent, useful output from your analysis.
从分析结果中产出连贯、实用的输出内容。
Step 1: Structure the Output
步骤1:构建输出结构
Choose appropriate format based on use case:
- Executive summary: Quick overview for decisions
- Full report: Comprehensive documentation
- Action items: Practical next steps
Use the template at assets/research-report-template.md.
根据使用场景选择合适的格式:
- 执行摘要:供决策使用的快速概述
- 完整报告:全面的研究记录
- 行动项:实用的后续步骤
使用[assets/research-report-template.md]中的模板。
Step 2: Write the Synthesis
步骤2:撰写整合内容
Key principles:
- Lead with most important findings
- Connect related concepts
- Note confidence levels
- Acknowledge limitations
- Cite sources
核心原则:
- 以最重要的研究发现开篇
- 关联相关概念
- 标注可信度等级
- 承认局限性
- 引用来源
Step 3: Include Actionable Elements
步骤3:包含可执行内容
End with practical outputs:
- Key takeaways (3-5 points)
- Recommendations
- Further research suggestions
- Decision points
结尾添加实用输出:
- 关键要点(3-5个)
- 建议
- 进一步研究的建议
- 决策节点
Complete Example
完整示例
Scenario: Research "Best practices for API versioning"
场景:研究"API版本控制的最佳实践"
Phase 1 - Planning:
Research Question: What are the best practices for API versioning?
Sub-questions:
1. What versioning strategies exist?
2. What are pros/cons of each?
3. What do major companies use?
4. What do experts recommend?
Search Plan:
- "API versioning strategies comparison"
- "REST API versioning best practices 2024"
- "API versioning header vs URL vs query parameter"
- "large companies API versioning approach"Phase 2 - Execution:
Query 1: "API versioning strategies comparison"
- Found: URL versioning, header versioning, query parameter
- Key insight: URL versioning most common, header more "RESTful"
- Sources: REST API tutorial, Martin Fowler blog
Query 2: "REST API versioning best practices 2024"
- Found: Semantic versioning principles apply
- Key insight: Version only when breaking changes
- Sources: API design guides, Stack Overflow discussionsPhase 3 - Analysis:
Consensus Points:
- Version only for breaking changes
- Be consistent within an API
- Document version lifecycle
Conflicts:
- URL vs header placement (no clear winner)
- When to deprecate old versions
Gaps:
- Limited data on performance impact
- Few studies on developer experiencePhase 4 - Synthesis:
Key Findings:
1. Three main strategies exist (URL, header, query param)
2. URL versioning is most common and discoverable
3. Header versioning is considered more "pure" REST
4. Version only on breaking changes
5. Major companies split between approaches
Recommendations:
- Use URL versioning for public APIs (discoverability)
- Consider header versioning for internal APIs
- Document deprecation timeline clearly
- Use semantic versioning principles阶段1 - 规划:
研究问题:API版本控制的最佳实践有哪些?
子问题:
1. 存在哪些版本控制策略?
2. 每种策略的优缺点是什么?
3. 大型企业使用哪种策略?
4. 专家推荐哪种策略?
搜索计划:
- "API versioning strategies comparison"
- "REST API versioning best practices 2024"
- "API versioning header vs URL vs query parameter"
- "large companies API versioning approach"阶段2 - 执行:
查询1:"API versioning strategies comparison"
- 发现:URL版本控制、请求头版本控制、查询参数版本控制
- 关键见解:URL版本控制最为常见,请求头版本控制更符合REST规范
- 来源:REST API教程、Martin Fowler博客
查询2:"REST API versioning best practices 2024"
- 发现:语义化版本控制原则适用
- 关键见解:仅在发生破坏性变更时才进行版本迭代
- 来源:API设计指南、Stack Overflow讨论阶段3 - 分析:
共识点:
- 仅在发生破坏性变更时才进行版本迭代
- 在同一个API中保持一致性
- 记录版本生命周期
冲突点:
- 版本放置在URL还是请求头(无明确最优解)
- 何时弃用旧版本
信息缺口:
- 性能影响的相关数据有限
- 关于开发者体验的研究较少阶段4 - 信息整合:
关键发现:
1. 存在三种主要策略(URL、请求头、查询参数)
2. URL版本控制最为常见且易于发现
3. 请求头版本控制被认为更"纯粹"的REST方式
4. 仅在发生破坏性变更时才进行版本迭代
5. 大型企业在策略选择上存在分歧
建议:
- 面向公开API使用URL版本控制(易于发现)
- 内部API可考虑使用请求头版本控制
- 清晰记录弃用时间线
- 遵循语义化版本控制原则Quality Checklist
质量检查清单
Before completing research, verify:
- Clear research questions were defined
- Multiple queries were executed (minimum 3-5)
- Sources were evaluated for credibility
- Findings are organized by theme
- Consensus and conflicts are noted
- Confidence levels are indicated
- Limitations are acknowledged
- Output is actionable
完成研究前,请验证:
- 已定义清晰的研究问题
- 已执行多个查询(至少3-5个)
- 已评估来源的可信度
- 已按主题整理研究发现
- 已记录共识与冲突点
- 已标注可信度等级
- 已承认局限性
- 输出内容具备可执行性
Reference Materials
参考资料
For detailed guidance, see:
- Research Methodology - In-depth methodology background
- Output Formats - Detailed format specifications
如需详细指导,请参阅:
- 研究方法 - 深入的方法学背景
- 输出格式 - 详细的格式规范
Templates
模板
- Research Plan Template
- Research Report Template
- Source Evaluation Checklist
- 研究计划模板
- 研究报告模板
- 来源评估清单
Limitations
局限性
This workflow has the following limitations:
- Quality depends on available web search capability
- Cannot access paywalled or restricted content
- Time-intensive for comprehensive research
- Synthesis quality depends on agent capabilities
- May miss very recent developments not yet indexed
该工作流存在以下局限性:
- 研究质量取决于可用的网页搜索能力
- 无法访问付费墙或受限内容
- 全面研究耗时较长
- 信息整合质量取决于Agent的能力
- 可能遗漏尚未被索引的最新发展
Related Skills
相关Skill
- web-search: For executing individual web searches (used within this workflow)
- web-search:用于执行单个网页搜索(在本工作流中使用)