report-writing

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Report Writing Skill

研究报告撰写Skill

This skill provides structured guidance for transforming research findings into well-organized, professional reports. It ensures consistency, clarity, and completeness across all research outputs.

本Skill为将研究成果转化为结构清晰、专业规范的报告提供结构化指导，确保所有研究产出内容一致、表述清晰且完整。

When to Use This Skill

何时使用本Skill

Invoke this skill when:

Finalizing Research: Converting raw research notes and findings into a formal deliverable
Creating Documentation: Producing technical documentation, white papers, or analysis reports
Synthesizing Multiple Sources: Combining insights from various sub-agent research tasks into a unified narrative
Stakeholder Communication: Preparing reports for executive review, technical teams, or external audiences
Knowledge Preservation: Documenting research methodology and findings for future reference

Do NOT use this skill for:

Quick summaries or informal notes (use simple markdown instead)
Real-time status updates (use TODO lists)
Raw data dumps (use structured data files)

调用本Skill的场景：

研究成果定稿：将原始研究笔记和成果转化为正式交付物
创建文档：编写技术文档、白皮书或分析报告
整合多源内容：将来自多个子Agent研究任务的见解整合为统一的叙述内容
利益相关方沟通：为高管评审、技术团队或外部受众准备报告
知识留存：记录研究方法和成果，供未来参考

请勿在以下场景使用本Skill：

快速摘要或非正式笔记（使用简单Markdown即可）
实时状态更新（使用待办事项列表）
原始数据转储（使用结构化数据文件）

Report Structure Template

报告结构模板

Every research report MUST follow this hierarchical structure. Adapt section depth based on report complexity.

所有研究报告必须遵循以下层级结构。可根据报告复杂度调整章节深度。

1. Executive Summary

1. 执行摘要

Purpose: Provide a standalone overview for readers who may not read the full report.

Contents:

Research objective (1-2 sentences)
Key findings (3-5 bullet points)
Primary recommendations or conclusions
Critical limitations or caveats

Length: 150-300 words (1 page maximum)

Writing Tip: Write this section LAST, after all other sections are complete.

目的：为可能不会阅读完整报告的读者提供独立的概述。

内容：

研究目标（1-2句话）
关键发现（3-5个要点）
主要建议或结论
关键限制或注意事项

篇幅：150-300字（最多1页）

写作技巧：最后撰写本章节，待所有其他章节完成后再进行。

2. Introduction/Background

2. 引言/背景

Purpose: Establish context and frame the research question.

Contents:

Problem statement or research question
Why this research matters (business/technical impact)
Scope boundaries (what IS and IS NOT covered)
Brief overview of approach taken

Length: 200-500 words

目的：建立研究背景，明确研究问题。

内容：

问题陈述或研究问题
本研究的重要性（业务/技术影响）
研究范围界定（包含与不包含的内容）
研究方法概述

篇幅：200-500字

3. Methodology

3. 研究方法

Purpose: Enable reproducibility and establish credibility.

Contents:

Data sources consulted (with dates accessed)
Search strategies and queries used
Selection criteria for sources
Tools and techniques employed
Limitations of the methodology

Example Format:

markdown

undefined

目的：确保研究可复现，提升研究可信度。

内容：

参考的数据来源（含访问日期）
使用的搜索策略和查询语句
数据源选择标准
采用的工具和技术
研究方法的局限性

示例格式：

markdown

undefined

Data Collection

数据收集

Primary sources: [List with access dates]
Search queries: [Exact queries used]
Time range: [Date boundaries for research]

主要来源：[列出来源及访问日期]
查询语句：[使用的精确查询词]
时间范围：[研究的日期边界]

Analysis Approach

分析方法

[Describe analytical framework]
[Note any tools or models used]

undefined

[描述分析框架]
[注明使用的工具或模型]

undefined

4. Findings

4. 研究发现

Purpose: Present discovered facts objectively, without interpretation.

Contents:

Organized by theme, source type, or chronology
Each finding clearly attributed to source
Quantitative data in tables/charts when applicable
Direct quotes for critical evidence

Structure Options:

Thematic: Group by topic or category
Comparative: Side-by-side analysis of alternatives
Chronological: Timeline of developments
Source-based: Organized by information source

目的：客观呈现发现的事实，不加入主观解读。

内容：

按主题、来源类型或时间顺序组织内容
每个发现都明确标注来源
定量数据尽可能以表格/图表呈现
关键证据使用直接引语

结构选项：

主题式：按主题或类别分组
对比式：对备选方案进行并列分析
时序式：按发展时间线呈现
来源式：按信息来源组织

5. Analysis

5. 分析解读

Purpose: Interpret findings and extract meaning.

Contents:

Patterns and trends identified
Contradictions or gaps in evidence
Implications of findings
Comparison with existing knowledge
Confidence levels for conclusions

Analysis Framework:

markdown

undefined

目的：解读研究发现，提炼核心意义。

内容：

识别出的模式和趋势
证据中的矛盾或空白
研究发现的潜在影响
与现有知识的对比
结论的置信度

分析框架：

markdown

undefined

Pattern Analysis

模式分析

[What recurring themes emerge?]

[呈现了哪些重复出现的主题？]

Gap Analysis

空白分析

[What questions remain unanswered?]

[哪些问题仍未得到解答？]

Confidence Assessment

置信度评估

High confidence: [Findings with strong evidence]
Medium confidence: [Findings with partial evidence]
Low confidence: [Tentative findings requiring validation]

undefined

高置信度：[有充分证据支持的发现]
中置信度：[有部分证据支持的发现]
低置信度：[初步发现，需进一步验证]

undefined

6. Conclusions

6. 结论

Purpose: Synthesize analysis into actionable insights.

Contents:

Direct answers to research questions
Prioritized recommendations (if applicable)
Suggested next steps or future research
Final assessment of confidence

Format:

markdown

undefined

目的：将分析结果整合为可落地的见解。

内容：

研究问题的直接答案
优先级排序的建议（如适用）
建议的后续步骤或未来研究方向
最终的置信度评估

格式：

markdown

undefined

Key Conclusions

核心结论

[Most important conclusion]
[Second conclusion]
[Third conclusion]

[最重要的结论]
[次要结论]
[第三结论]

Recommendations

建议

[Priority 1 action item]
[Priority 2 action item]

[优先级1行动项]
[优先级2行动项]

Future Research Directions

未来研究方向

[Unanswered questions to explore]

undefined

[待探索的未解答问题]

undefined

7. References

7. 参考文献

Purpose: Enable verification and further exploration.

Contents:

All sources cited in the report
URLs with access dates
Proper attribution for all quoted material

目的：方便验证和进一步探索。

内容：

报告中引用的所有来源
含访问日期的URL
所有引述内容的正确署名

Citation Formatting Guidelines

引用格式指南

In-Text Citations

文内引用

Use numbered references in square brackets for inline citations:

markdown

Recent studies indicate a 40% improvement in efficiency [1]. This aligns with
earlier findings on automation benefits [2, 3].

For direct quotes, include page numbers or section identifiers:

markdown

According to the official documentation, "the system supports up to 10,000
concurrent connections" [4, Section 3.2].

使用方括号包裹的编号作为行内引用：

markdown

近期研究表明效率提升了40% [1]。这与之前关于自动化收益的研究结果一致 [2, 3]。

直接引语需包含页码或章节标识符：

markdown

根据官方文档，“系统支持最多10000个并发连接” [4, 第3.2节]。

Reference List Format

参考文献列表格式

Use a consistent format for all references:

Web Sources:

markdown

[1] Author/Organization. "Article Title." Website Name. URL. Accessed: YYYY-MM-DD.

Academic Papers:

markdown

[2] Author(s). "Paper Title." Journal/Conference Name, Year. DOI/URL.

Documentation:

markdown

[3] "Document Title." Product Name Documentation, Version X.X. URL. Accessed: YYYY-MM-DD.

News Articles:

markdown

[4] Author. "Headline." Publication Name, Date Published. URL.

所有参考文献需使用统一格式：

网络来源：

markdown

[1] 作者/机构。“文章标题”。网站名称。URL。访问日期：YYYY-MM-DD。

学术论文：

markdown

[2] 作者。“论文标题”。期刊/会议名称，年份。DOI/URL。

技术文档：

markdown

[3] “文档标题”。产品名称文档，版本X.X。URL。访问日期：YYYY-MM-DD。

新闻文章：

markdown

[4] 作者。“标题”。出版物名称，发布日期。URL。

Citation Best Practices

引用最佳实践

Always include access dates for web sources (content may change)
Prefer primary sources over secondary reports
Note version numbers for software documentation
Archive volatile sources when possible (use archive.org links)
Verify link validity before finalizing report

网络来源务必包含访问日期（内容可能变更）
优先使用原始来源而非二次报告
软件文档需注明版本号
尽可能存档易变来源（使用archive.org链接）
定稿前验证链接有效性

Writing Style Recommendations

写作风格建议

Style Selection Guide

风格选择指南

Audience	Style	Characteristics
Executives	Executive	Concise, outcome-focused, minimal jargon
Technical Teams	Technical	Detailed, precise terminology, includes code/data
Academic/Research	Academic	Formal, extensive citations, methodological rigor
General Stakeholders	Balanced	Clear explanations, moderate detail, accessible

受众群体	风格类型	特点
高管	高管风格	简洁、以结果为导向、专业术语极少
技术团队	技术风格	内容详尽、术语精准、包含代码/数据
学术/研究人员	学术风格	正式规范、引用丰富、方法严谨
普通利益相关方	平衡风格	解释清晰、细节适度、易于理解

Executive Style

高管风格

Characteristics:

Lead with conclusions and recommendations
Use bullet points liberally
Limit technical jargon; define necessary terms
Focus on business impact and ROI
Keep paragraphs short (3-4 sentences max)

Example:

markdown

undefined

特点：

以结论和建议开篇
大量使用要点列表
尽量减少专业术语；必要术语需定义
聚焦业务影响和投资回报率（ROI）
段落简短（最多3-4句话）

示例：

markdown

undefined

Key Finding: Cloud Migration Reduces Costs by 35%

关键发现：云迁移降低35%成本

Bottom Line: Migrating to cloud infrastructure will reduce operational costs by $2.4M annually while improving system reliability.

Recommended Action: Approve Phase 1 migration by Q2 2025.

Risk Level: Low - Similar migrations have 94% success rate.

undefined

核心结论：迁移到云基础设施每年将减少240万美元运营成本，同时提升系统可靠性。

建议行动：在2025年第二季度前批准第一阶段迁移。

风险等级：低 - 类似迁移的成功率达94%。

undefined

Technical Style

技术风格

Characteristics:

Include implementation details
Use precise technical terminology
Provide code samples, configurations, or specifications
Document edge cases and limitations
Include performance metrics and benchmarks

Example:

markdown

undefined

特点：

包含实现细节
使用精准的专业术语
提供代码示例、配置或规格说明
记录边缘情况和局限性
包含性能指标和基准数据

示例：

markdown

undefined

Implementation: Rate Limiting Configuration

实现：限流配置

The API gateway implements token bucket rate limiting with the following parameters:

Parameter	Value	Rationale
Bucket Size	1000	Handles burst traffic
Refill Rate	100/sec	Sustainable throughput
Key Strategy	IP + User ID	Prevents abuse while supporting legitimate use

python

rate_limiter = TokenBucket(
    capacity=1000,
    refill_rate=100,
    key_func=lambda req: f"{req.ip}:{req.user_id}"
)

undefined

API网关采用令牌桶限流算法，参数如下：

参数	值	依据
令牌桶容量	1000	应对突发流量
refill速率	100/秒	可持续吞吐量
密钥策略	IP + 用户ID	防止滥用同时支持合法使用

python

rate_limiter = TokenBucket(
    capacity=1000,
    refill_rate=100,
    key_func=lambda req: f"{req.ip}:{req.user_id}"
)

undefined

Academic Style

学术风格

Characteristics:

Formal third-person voice
Extensive literature review
Detailed methodology documentation
Statistical rigor where applicable
Acknowledge limitations explicitly

Example:

markdown

undefined

特点：

使用正式的第三人称语气
包含详尽的文献综述
详细记录研究方法
适用场景下保持统计严谨性
明确承认研究局限性

示例：

markdown

undefined

Literature Review

文献综述

Previous research in automated code review systems has demonstrated significant potential for defect detection. Smith et al. (2023) reported a 23% reduction in production defects when implementing static analysis tools [1]. However, Johnson and Lee (2024) note that these gains are contingent upon proper configuration and team adoption [2].

The present study extends this work by examining the integration of large language models into the review pipeline, an approach not addressed in prior literature.

undefined

此前关于自动化代码审查系统的研究已证明其在缺陷检测方面的巨大潜力。Smith等人（2023）报告称，实施静态分析工具后生产缺陷减少了23% [1]。然而，Johnson和Lee（2024）指出，这些收益取决于正确的配置和团队的接受度 [2]。

本研究拓展了现有成果，探讨了将大语言模型整合到审查流程中的方法，这是现有文献未涉及的方向。

undefined

General Guidelines (All Styles)

通用指南（所有风格适用）

Active voice preferred: "The team implemented" not "It was implemented by the team"
Specific over vague: "37% increase" not "significant increase"
Present tense for findings: "The data shows" not "The data showed"
Consistent terminology: Choose one term and use it throughout
Avoid hedging excess: Limit "may," "might," "could possibly"

优先使用主动语态：“团队实施了...”而非“...被团队实施”
具体表述优于模糊描述：“增长37%”而非“显著增长”
研究发现使用一般现在时：“数据显示...”而非“数据曾显示...”
术语保持一致：选择一个术语并全程使用
避免过度模糊表述：限制“可能”“也许”“或许”等词的使用

Quality Checklist Before Submission

提交前质量检查清单

Structure Verification

结构验证

Content Quality

内容质量

Citation Integrity

引用完整性

Writing Quality

写作质量

Formatting

格式规范

Final Review

最终审查

Examples of Report Sections

报告章节示例

Example: Executive Summary

示例：执行摘要

markdown

undefined

markdown

undefined

Executive Summary

执行摘要

This report evaluates three cloud database solutions for the customer analytics platform migration: AWS Aurora, Google Cloud Spanner, and Azure Cosmos DB.

Key Findings:

AWS Aurora offers the lowest total cost of ownership ($145K/year)
Google Cloud Spanner provides superior global consistency guarantees
Azure Cosmos DB integrates best with existing Microsoft infrastructure
All three solutions meet performance requirements (< 50ms p99 latency)

Recommendation: Proceed with AWS Aurora for Phase 1, with architecture designed to allow future multi-cloud expansion.

Timeline: Implementation achievable within Q2 2025 with existing team.

Confidence Level: High - Based on proof-of-concept testing and vendor consultations.

undefined

本报告评估了客户分析平台迁移的三款云数据库解决方案：AWS Aurora、Google Cloud Spanner和Azure Cosmos DB。

关键发现：

AWS Aurora拥有最低的总拥有成本（每年14.5万美元）
Google Cloud Spanner提供更出色的全局一致性保障
Azure Cosmos DB与现有微软基础设施集成度最高
三款解决方案均满足性能要求（P99延迟<50ms）

建议：推进AWS Aurora作为第一阶段方案，架构设计需支持未来多云扩展。

时间线：现有团队可在2025年第二季度内完成实施。

置信度：高 - 基于概念验证测试和厂商咨询结果。

undefined

Example: Methodology Section

示例：研究方法章节

markdown

undefined

markdown

undefined

Methodology

研究方法

Research Approach

研究途径

This analysis employed a mixed-methods approach combining:

Vendor documentation review
Technical proof-of-concept testing
Industry analyst report analysis
Peer organization interviews

本分析采用混合方法，结合：

厂商文档审查
技术概念验证测试
行业分析师报告分析
同行企业访谈

Data Sources

数据来源

Source Type	Sources Consulted	Date Range
Vendor Docs	AWS, GCP, Azure official documentation	Dec 2024
Analyst Reports	Gartner, Forrester database evaluations	2024
Technical Tests	Internal POC environment	Dec 15-22, 2024
Interviews	3 peer organizations (anonymized)	Dec 2024

来源类型	参考来源	时间范围
厂商文档	AWS、GCP、Azure官方文档	2024年12月
分析师报告	Gartner、Forrester数据库评估	2024年
技术测试	内部概念验证环境	2024年12月15-22日
访谈	3家同行企业（匿名）	2024年12月

Evaluation Criteria

评估标准

Solutions were evaluated against weighted criteria:

Performance (30%): Latency, throughput, scalability
Cost (25%): TCO over 3 years including migration
Reliability (20%): SLA guarantees, disaster recovery
Integration (15%): Compatibility with existing stack
Vendor Support (10%): Documentation, support quality

解决方案基于加权标准进行评估：

性能（30%）：延迟、吞吐量、可扩展性
成本（25%）：3年总拥有成本，含迁移费用
可靠性（20%）：SLA保障、灾难恢复
集成度（15%）：与现有技术栈的兼容性
厂商支持（10%）：文档质量、支持服务水平

Limitations

局限性

POC testing limited to 72-hour duration
Cost projections based on current pricing (subject to change)
Interview sample size limits generalizability

undefined

概念验证测试仅持续72小时
成本预测基于当前定价（可能变动）
访谈样本量有限，结论推广性受限

undefined

Example: Findings Section

示例：研究发现章节

markdown

undefined

markdown

undefined

Findings

研究发现

Performance Comparison

性能对比

All three solutions demonstrated acceptable performance for the target workload of 10,000 queries per second:

Solution	Avg Latency	P99 Latency	Max Throughput
AWS Aurora	12ms	45ms	15,000 QPS
Cloud Spanner	15ms	42ms	18,000 QPS
Cosmos DB	18ms	48ms	12,000 QPS

Source: Internal POC testing, December 2024 [1]

三款解决方案均满足目标工作负载（每秒10000次查询）的性能要求：

解决方案	平均延迟	P99延迟	最大吞吐量
AWS Aurora	12ms	45ms	15000 QPS
Cloud Spanner	15ms	42ms	18000 QPS
Cosmos DB	18ms	48ms	12000 QPS

来源：内部概念验证测试，2024年12月 [1]

Cost Analysis

成本分析

Three-year total cost of ownership analysis:

AWS Aurora: $435,000

Compute: $180,000
Storage: $95,000
Data transfer: $85,000
Support: $75,000

Google Cloud Spanner: $520,000

[Detailed breakdown...]

Azure Cosmos DB: $485,000

[Detailed breakdown...]

Source: Vendor pricing calculators and enterprise discount estimates [2, 3, 4]

undefined

三年总拥有成本分析：

AWS Aurora：43.5万美元

计算资源：18万美元
存储：9.5万美元
数据传输：8.5万美元
支持服务：7.5万美元

Google Cloud Spanner：52万美元

[详细分项...]

Azure Cosmos DB：48.5万美元

[详细分项...]

来源：厂商定价计算器和企业折扣估算 [2, 3, 4]

undefined

Example: Analysis Section

示例：分析解读章节

markdown

undefined

markdown

undefined

Analysis

分析解读

Cost-Performance Trade-offs

成本-性能权衡

While AWS Aurora offers the lowest TCO, Cloud Spanner's 20% higher cost delivers measurably better global consistency. For applications requiring strong consistency across regions, this premium may be justified.

The cost difference primarily stems from:

Cloud Spanner's TrueTime infrastructure overhead
AWS Aurora's more aggressive reserved instance discounts
Different approaches to cross-region replication

虽然AWS Aurora拥有最低的总拥有成本，但Cloud Spanner高出20%的成本带来了显著更优的全局一致性。对于需要跨区域强一致性的应用场景，这一溢价是合理的。

成本差异主要源于：

Cloud Spanner的TrueTime基础设施开销
AWS Aurora更优惠的预留实例折扣
跨区域复制的不同实现方式

Risk Assessment

风险评估

Risk	Likelihood	Impact	Mitigation
Vendor lock-in	High	Medium	Abstract data layer
Price increases	Medium	Medium	3-year commitment
Service outage	Low	High	Multi-region deployment

风险	可能性	影响	缓解措施
厂商锁定	高	中	抽象数据层
价格上涨	中	中	3年承诺协议
服务中断	低	高	多区域部署

Confidence Assessment

置信度评估

High Confidence:

Performance meets requirements (validated via POC)
AWS Aurora is most cost-effective option

Medium Confidence:

3-year cost projections (pricing may change)
Integration complexity estimates

Low Confidence:

Long-term vendor roadmap alignment

undefined

高置信度：

性能满足要求（已通过概念验证验证）
AWS Aurora是最具成本效益的选项

中置信度：

3年成本预测（定价可能变动）
集成复杂度估算

低置信度：

长期厂商路线图对齐情况

undefined

Example: References Section

示例：参考文献章节

markdown

undefined

markdown

undefined

References

参考文献

[1] Internal Engineering Team. "Database POC Test Results." Internal Documentation. December 22, 2024.

[2] Amazon Web Services. "Amazon Aurora Pricing." AWS Documentation. https://aws.amazon.com/aurora/pricing/. Accessed: December 20, 2024.

[3] Google Cloud. "Cloud Spanner Pricing." Google Cloud Documentation. https://cloud.google.com/spanner/pricing. Accessed: December 20, 2024.

[4] Microsoft Azure. "Azure Cosmos DB Pricing." Azure Documentation. https://azure.microsoft.com/pricing/details/cosmos-db/. Accessed: December 20, 2024.

[5] Gartner. "Magic Quadrant for Cloud Database Management Systems." Gartner Research, November 2024. (Subscription required)

[6] Smith, J. and Chen, L. "Comparative Analysis of Distributed Databases." Proceedings of VLDB 2024. DOI: 10.14778/example.

---

[1] 内部工程团队。“数据库概念验证测试结果”。内部文档。2024年12月22日。

[2] Amazon Web Services。“Amazon Aurora定价”。AWS文档。 https://aws.amazon.com/aurora/pricing/。访问日期：2024年12月20日。

[3] Google Cloud。“Cloud Spanner定价”。Google Cloud文档。 https://cloud.google.com/spanner/pricing。访问日期：2024年12月20日。

[4] Microsoft Azure。“Azure Cosmos DB定价”。Azure文档。 https://azure.microsoft.com/pricing/details/cosmos-db/。访问日期：2024年12月20日。

[5] Gartner。“云数据库管理系统魔力象限”。 Gartner研究报告，2024年11月。（需订阅）

[6] Smith, J. 和 Chen, L. “分布式数据库对比分析”。 VLDB 2024会议论文集。DOI: 10.14778/example.

---

Integration with Research Workflow

与研究工作流的整合

This skill integrates with the broader research workflow as follows:

Research Request → Data Collection → Analysis → [REPORT WRITING] → Verification → Delivery
                                                      ↑
                                                 This Skill

Inputs Expected:

Completed research findings (from sub-agents or direct research)
Original research request/questions
Source materials and citations
Any constraints (length, audience, format)

Outputs Produced:

Formatted report following structure template
Complete reference list
Executive summary for quick consumption

Quality Gates:

Report must pass quality checklist before marking complete
All citations must be verifiable
Conclusions must trace back to evidence in findings

本Skill与更广泛的研究工作流整合方式如下：

研究请求 → 数据收集 → 分析 → [REPORT WRITING] → 验证 → 交付
                                                      ↑
                                                 本Skill

预期输入：

已完成的研究发现（来自子Agent或直接研究）
原始研究请求/问题
来源材料和引用信息
任何约束条件（篇幅、受众、格式）

产出内容：

遵循结构模板的格式化报告
完整的参考文献列表
供快速浏览的执行摘要

质量关卡：

报告必须通过质量检查清单才可标记完成
所有引用必须可验证
结论必须可追溯到研究发现中的证据