deep-research
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseDeep Research
深度研究
Produce Gemini Deep Research-quality output: rich artifacts with embedded screenshots, mermaid diagrams, comparison tables, and narrative synthesis. Tuned for developer decisions — framework selection, architecture patterns, dependency evaluation, competitive analysis.
生成具备Gemini深度研究级别的成果:包含嵌入式截图、Mermaid图表、对比表格及叙事性综合分析的丰富产出物。专为开发者决策场景优化——如框架选型、架构模式、依赖评估、竞品分析等。
When to Use This Skill
何时使用该Skill
- "Research the current state of X"
- "Compare Framework A vs Framework B"
- "What are the best approaches for..."
- "Deep dive into..."
- Any request where the answer requires synthesizing information from many sources
Do NOT use for: quick factual lookups, single-source answers, or "find me a CSS button" (use instead).
design-lookup- 「调研X领域的当前现状」
- 「对比框架A与框架B」
- 「解决...问题的最佳方案有哪些」
- 「深入探索...」
- 任何需要整合多信息源内容的请求
请勿用于:快速事实查询、单信息源即可解答的问题,或「帮我找一个CSS按钮」(请使用替代)。
design-lookupInput Protocol — Before Any Search
输入协议 —— 搜索前准备
- Decompose the topic into 3-5 research axes.
- Example: "Compare Next.js vs Remix" → Performance, DX, Ecosystem, Deployment, Community
- Identify the decision context — what is the user actually deciding?
- Framework choice? Architecture pattern? Build vs buy? Migration risk?
- Draft a research plan — present 3-5 axes with planned queries to the user.
- Save it as an artifact (e.g., ).
research_plan.md - Proceed on approval, or refine if the user redirects scope.
- Save it as an artifact (e.g.,
- 主题拆解:将研究主题拆解为3-5个研究维度。
- 示例:「对比Next.js与Remix」→ 性能、DX(开发者体验)、生态系统、部署、社区
- 明确决策场景:用户实际要做什么决策?
- 框架选型?架构模式选择?自研还是采购?迁移风险评估?
- 制定研究计划:向用户展示3-5个研究维度及对应的拟用查询词。
- 将计划保存为产出物(如)。
research_plan.md - 获得用户批准后再推进,若用户调整范围则相应优化计划。
- 将计划保存为产出物(如
Phase 1: Breadth Scan
第一阶段:广度扫描
Goal: Map the landscape. Find what exists before reading anything.
- Run 5-8 parallel searches across different axes. Use at least two tools:
- — broad topic queries
tavily_search - — alternate search perspective
search_web - — delegate an entire sub-question (powerful for "state of X" queries)
tavily_research
- Dev-specific breadth:
- or
search_code— find relevant GitHub repossearch_repositories - Search npm trends, bundle sizes, download counts when evaluating packages
- Search for migration stories: "migrating from X to Y" experience reports
- Collect 15-25 candidate URLs, not 5. Score each by authority tier (see references/research-heuristics.md).
- Do not stop at snippets. Snippets are for candidate selection only.
Output: Candidate source list with tier ratings. Present to user if interactive, or proceed if autonomous.
目标:梳理研究领域全貌,先明确有哪些信息源再深入阅读。
- 针对不同维度运行5-8个并行搜索,至少使用两种工具:
- —— 宽泛主题查询
tavily_search - —— 补充搜索视角
search_web - —— 委托处理整个子问题(对「X领域现状」类查询尤为高效)
tavily_research
- 开发者专属广度调研:
- 或
search_code—— 查找相关GitHub仓库search_repositories - 评估包时,搜索npm趋势、包体积、下载量数据
- 搜索迁移实践案例:「从X迁移到Y」的经验分享
- 收集15-25个候选URL,而非5个。根据权威等级为每个URL评分(参考references/research-heuristics.md)。
- 不要仅停留在搜索摘要:摘要仅用于候选信息源筛选。
产出:带权威等级评分的候选信息源列表。若为交互式场景则展示给用户,若为自主模式则直接推进。
Phase 2: Deep Read
第二阶段:深度阅读
Goal: Extract actual content — implementation details, code examples, benchmarks, data.
- Select the top 8-12 sources from Phase 1 (prioritize S and A tier).
- Full extraction — get the complete page content:
- or
tavily_extractfor text-heavy pagesread_url_content - to follow documentation multi-page structures
tavily_crawl - to screenshot key pages (UIs, dashboards, architecture diagrams)
browser_subagent - (GitHub MCP) to read actual source code from repos
get_file_contents
- Analyze each source:
- Extract specific claims, numbers, patterns, code examples
- Note the authority tier and any bias (is this the framework's own marketing?)
- Tag findings by research axis
- Self-correction: If a source is fluff (marketing-only, thin tutorial, SEO filler):
- Discard it
- Run a refined follow-up search with more specific terms
- Try adding: "benchmark", "technical deep dive", "lessons learned", "postmortem"
Output: Annotated source notes organized by axis.
目标:提取实际内容——实现细节、代码示例、基准测试数据等。
- 从第一阶段的候选中筛选出排名前8-12的信息源(优先选择S级和A级)。
- 完整内容提取:获取页面全部内容:
- 针对文本密集型页面,使用或
tavily_extractread_url_content - 使用处理多页文档结构
tavily_crawl - 使用对关键页面(UI、仪表盘、架构图)截图
browser_subagent - 使用(GitHub MCP)读取仓库中的实际源代码
get_file_contents
- 针对文本密集型页面,使用
- 逐个分析信息源:
- 提取具体结论、数据、模式、代码示例
- 记录信息源的权威等级及可能存在的 bias(是否为框架官方营销内容?)
- 按研究维度为研究发现打标签
- 自我修正:若信息源内容空洞(仅营销话术、浅显教程、SEO填充内容):
- 丢弃该信息源
- 使用更精准的关键词进行补充搜索
- 尝试添加:「benchmark」「technical deep dive」「lessons learned」「postmortem」等关键词
产出:按研究维度整理的带注释信息源笔记。
Phase 3: Synthesis
第三阶段:内容整合
Goal: Build the research briefing artifact. This is the main deliverable.
- Choose the report template from references/report-templates.md:
- Comprehensive Brief — for landscape/state-of-the-art research
- Comparison Brief — for head-to-head evaluations
- Write the report as a rich markdown artifact:
- Narrative prose in the executive summary — not bullets, not lists. Write as if briefing a tech lead.
- Comparison tables with real data extracted from sources
- Mermaid diagrams for architecture, decision trees, ecosystem maps
- Embedded screenshots captured via during Phase 2
browser_subagent - Code examples pulled from actual repos or docs
- Use for custom visualizations when no screenshot captures the concept
generate_image
- Cite every claim — link to the source URL inline. Use the format: .
[Source Name](URL) - Gap analysis — explicitly call out:
- What couldn't be determined and why
- Conflicting information between sources
- Areas where only low-tier sources were found
Output: The research artifact (e.g., ).
research_report.md目标:生成研究简报成果,这是核心交付物。
- 从references/report-templates.md中选择合适的报告模板:
- 综合简报 —— 适用于领域全景/前沿技术调研
- 对比简报 —— 适用于一对一评估
- 撰写富格式Markdown报告:
- 执行摘要采用叙事性 prose,而非项目符号列表。以向技术负责人汇报的口吻撰写。
- 加入从信息源提取的真实数据制作对比表格
- 使用Mermaid图表展示架构、决策树、生态系统图谱
- 嵌入第二阶段通过捕获的截图
browser_subagent - 从实际仓库或文档中提取代码示例
- 若截图无法呈现概念,使用生成自定义可视化内容
generate_image
- 为所有结论添加引用 —— 内联链接至信息源URL。使用格式:。
[信息源名称](URL) - 差距分析:明确指出:
- 无法确定的内容及原因
- 不同信息源之间的矛盾信息
- 仅能找到低等级信息源的领域
产出:研究成果物(如)。
research_report.mdPhase 4: Iteration
第四阶段:迭代优化
Goal: Fill gaps identified in Phase 3.
- Review the gap analysis section of your report.
- For each fillable gap:
- Run 1-2 targeted searches with refined queries
- Extract and read the results
- Update the report artifact in-place
- Max 3 total iterations (Phase 1-3 = round 1, then up to 2 more targeted rounds).
- After final iteration, mark remaining gaps as "Unresolved" with explanation.
目标:填补第三阶段中发现的信息空白。
- 审阅报告中的差距分析部分。
- 针对每个可填补的空白:
- 运行1-2次精准搜索
- 提取并阅读结果
- 直接更新报告成果物
- 最多进行3轮迭代(第一至第三阶段为第1轮,之后可进行最多2轮针对性补充)。
- 最终迭代后,将剩余空白标记为「未解决」并说明原因。
Tool Strategy
工具策略
| Purpose | Primary | Fallback |
|---|---|---|
| Topic discovery | | |
| Delegated deep research | | Manual multi-search |
| Full page extraction | | |
| Multi-page docs | | |
| Visual evidence | | |
| GitHub analysis | | |
| Architecture diagrams | Mermaid in markdown | |
| Data visualization | Markdown tables | |
| 用途 | 首选工具 | 备选工具 |
|---|---|---|
| 主题探索 | | |
| 委托式深度研究 | | 手动多工具搜索 |
| 整页内容提取 | | |
| 多页文档处理 | | |
| 视觉证据收集 | | |
| GitHub分析 | | 读取GitHub原始页面的 |
| 架构图制作 | Markdown中的Mermaid | |
| 数据可视化 | Markdown表格 | |
Quality Gates
质量校验清单
Before delivering the report, verify:
- Source diversity — at least 1 S-tier and 2 A-tier sources cited (or explicitly flagged as unavailable)
- Visual richness — at least 1 screenshot/image AND 1 diagram/table embedded
- Narrative quality — executive summary reads as prose, not bullet points
- Citation completeness — every factual claim links to a source
- Gap transparency — gaps and conflicts are explicitly documented
- Actionable output — recommendations section exists with ranked, specific advice
交付报告前,请验证:
- 信息源多样性 —— 至少引用1个S级和2个A级信息源(若无法获取则需明确标记)
- 视觉丰富度 —— 至少嵌入1张截图/图片 及 1个图表/表格
- 叙事质量 —— 执行摘要为 prose 格式,而非项目符号列表
- 引用完整性 —— 所有事实性结论均链接至信息源
- 差距透明度 —— 明确记录信息空白及矛盾内容
- 可行动性 —— 包含建议部分,提供分级的具体建议
Anti-Patterns
反模式
- Snippet-only research — stopping at search result descriptions without full extraction
- Text-wall reports — no visuals, no tables, no diagrams. The whole point is richness.
- Source-by-source organization — findings must be grouped thematically by research axis, not by URL
- Single-tool reliance — use at least 2 different search/extraction tools for source diversity
- Uncited claims — every substantive finding must link to its source
- Marketing echo — repeating a framework's own marketing claims without independent verification
- Premature stopping — delivering after 3-5 sources when the topic warrants 15+
- 仅依赖搜索摘要的研究:未提取完整内容就停留在搜索结果描述
- 纯文本报告:无视觉元素、表格或图表。本技能的核心就是产出丰富的成果
- 按信息源组织内容:研究发现需按研究维度主题分组,而非按URL分组
- 单一工具依赖:至少使用2种不同的搜索/提取工具以保证信息源多样性
- 无引用结论:所有实质性研究发现均需链接至其信息源
- 营销话术重复:未经过独立验证就直接引用框架官方营销内容
- 提前终止研究:当主题需要15+信息源时,仅调研3-5个就交付成果
References
参考资料
- Source authority scoring and query patterns: references/research-heuristics.md
- Report structure templates: references/report-templates.md
- 信息源权威评分及查询模式:references/research-heuristics.md
- 报告结构模板:references/report-templates.md