Back to Details

research

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Research — Hybrid Router + Fallback

Research — 混合路由器 + 后备工作流

The runtime orchestrator for the research domain. Architecture C: deterministic classification → specialist delegation OR own plan-decompose-search-synthesize-cite workflow.
研究领域的运行时编排器。架构C：确定性分类 → 专业技能委托 或 自有「规划-分解-搜索-合成-引用」工作流。

Portability

可移植性

Requires 
WebSearch
+ 
WebFetch
for the fallback workflow; specialist skills (
pulse
, 
grants
, 
litreview
, 
syllabus
, 
patent
, 
dossier
) must be present for delegation to work. Node.js with 
docx
package required if Q2 = document mode. Works in Claude Code CLI natively. In Claude.ai with web tools + Code Execution, the workflow is supported.
后备工作流需要
WebSearch
+ 
WebFetch
；委托功能需要专业技能（
pulse
、
grants
、
litreview
、
syllabus
、
patent
、
dossier
）已部署。若用户选择文档模式（Q2），则需要安装
docx
包的Node.js环境。原生支持Claude Code CLI。在配备Web工具+代码执行功能的Claude.ai中，该工作流同样受支持。

Distinct From 
engineering/autoresearch-agent

与
engineering/autoresearch-agent
的区别

These two skills share the word "research" but serve completely different use cases:
  • research/research/
     (this skill) — research-query router + fallback workflow ("Research X")
  • engineering/autoresearch-agent/
     — Karpathy's autonomous file-optimization experiment loop ("Make this code faster")
No overlap. They coexist.
这两个技能都包含「research」一词，但服务于完全不同的使用场景：
  • research/research/
    （当前技能）——研究查询路由器 + 后备工作流（「Research X」）
  • engineering/autoresearch-agent/
    ——Karpathy的自主文件优化实验循环（「Make this code faster」）
两者无重叠，可共存。

Hybrid Architecture (C)

混合架构（C）

Every invocation produces one of three outcomes:
  1. Delegation — Classified as specialist-domain. Routes there. User sees the specialist's output.
  2. Fallback execution — Classified as general research. Runs own plan → search → synthesize workflow.
  3. Clarification request — Classification ambiguous. Asks one forcing question to disambiguate, then routes.
The skill never silently runs its fallback when a specialist would have done better. Routing transparency is what makes the hybrid architecture trustworthy.
每次调用会产生以下三种结果之一：
  1. 委托——归类为专业领域，路由至对应技能，用户将看到专业技能的输出。
  2. 后备执行——归类为通用研究，运行自有「规划→搜索→合成」工作流。
  3. 请求澄清——分类模糊，提出一个明确问题以消除歧义，再进行路由。
当专业技能能提供更优结果时，该技能绝不会静默运行后备工作流。路由透明度是混合架构值得信赖的核心原因。

Specialist Registry

专业技能注册表

SpecialistRouting signalsDomain
pulse
reddit / hn / x / buzz / sentiment / trending / "what's people saying" / "pulse on" / "take the pulse" / "current conversation"Multi-source recency research
grants
NIH / grant / R01 / K-award / RePORTER / NOSI / "grants for" / FDA / "study section" / "principal investigator"NIH grant-funding intelligence
litreview
literature review / PICO / SPIDER / systematic review / "review papers on" / meta-analysisAcademic literature orientation
syllabus
syllabus / course outline / curriculum / "reading list" / "for my class" / "for my students"Course supplementary reading
patent
prior art / FTO / freedom to operate / patent / "patent landscape" / invention / novelty search / "ip landscape"Patent prior-art + landscape
dossier
"dossier on" / "due diligence" / "background check" / "prep me for" / "competitor research" / "investor diligence" / "interview prep" / "background on"Decision-grade entity research
专业技能路由信号领域
pulse
reddit / hn / x / buzz / sentiment / trending / "what's people saying" / "pulse on" / "take the pulse" / "current conversation"多源时效性研究
grants
NIH / grant / R01 / K-award / RePORTER / NOSI / "grants for" / FDA / "study section" / "principal investigator"NIH资助情报分析
litreview
literature review / PICO / SPIDER / systematic review / "review papers on" / meta-analysis学术文献导向研究
syllabus
syllabus / course outline / curriculum / "reading list" / "for my class" / "for my students"课程补充阅读资料整理
patent
prior art / FTO / freedom to operate / patent / "patent landscape" / invention / novelty search / "ip landscape"专利现有技术+格局分析
dossier
"dossier on" / "due diligence" / "background check" / "prep me for" / "competitor research" / "investor diligence" / "interview prep" / "background on"决策级实体研究

Agent Integrity Rules

Agent完整性规则

This skill obeys the research-pack convention:
  • Execution discipline (fallback only): Sequential searches. 1 q/sec rate limit. Confirm response received before next call.
  • Source discipline: Cite only sources returned by this session's tool calls. Training knowledge labeled 
    [Background — not from search]
    and excluded from counts.
  • Three-count tracking (fallback only): Queries sent / sources received / sources cited.
  • Retry policy: On failure → wait 3s → retry once → log. After 3 consecutive failures: stop, alert user.
  • Plan-tier detection: If delegated to Consensus-using specialist, that specialist handles detection. In fallback mode, surface any rate-limit signals.
  • Routing discipline: Never delegate silently. Always state the decision + accept override.
该技能遵循研究包约定：
  • 执行规范（仅后备工作流）：顺序搜索，每秒1次查询限制，确认收到响应后再发起下一次调用。
  • 来源规范：仅引用本次会话工具调用返回的来源。训练知识标记为
    [Background — not from search]
    ，不计入引用统计。
  • 三项计数跟踪（仅后备工作流）：已发送查询数 / 已接收来源数 / 已引用来源数。
  • 重试策略：失败后→等待3秒→重试一次→记录日志。连续3次失败后：停止执行，提醒用户。
  • 计划层级检测：若委托给使用Consensus的专业技能，由该技能负责检测。后备模式下，需显示任何速率限制信号。
  • 路由规范：绝不静默委托，始终说明决策并接受用户覆盖。

Phase 1: Grill-Me Intake (2–4 Questions)

阶段1：精简式导入（2–4个问题）

Intake is intentionally minimal — the goal is to route fast, not to interrogate. One question per turn.
导入流程刻意简化——目标是快速路由，而非过度询问。每轮仅提一个问题。

Q1 (always) — Research question

Q1（必问）——研究问题

What's the research question? State it in 1–2 sentences. Specific is better than broad — "AI for healthcare" gets you a vague survey; "How are health systems integrating LLM-based clinical decision support in 2026?" gets you a useful answer.
Why I'm asking: Specificity dictates classification accuracy and search precision. A vague question routes to fallback; a specific question often matches a specialist cleanly.
Refuse mush. If user says "research AI", push back once: "What about AI specifically — adoption, safety, capability, funding, regulation, comparison? Pick an angle."
你的研究问题是什么？用1-2句话表述。越具体越好——「AI for healthcare」会得到模糊的综述；「2026年医疗系统如何整合基于LLM的临床决策支持？」会得到有用的答案。
提问原因：问题的具体性决定了分类准确性和搜索精度。模糊问题会路由至后备工作流；具体问题通常能清晰匹配专业技能。
拒绝模糊表述。若用户仅说「research AI」，需提醒一次：「具体是关于AI的哪方面——落地应用、安全性、能力、资助、监管还是竞品对比？请选择一个方向。」

Q2 (always) — Output preference

Q2（必问）——输出偏好

What output do you want? Pick one:
  1. Quick chat briefing (5-min read, markdown in chat)
  2. Standalone document (.docx with citations, shareable)
Why I'm asking: Document mode triggers deeper search budgets and full audit logs. Chat mode optimizes for fast delivery.
Forcing choice.
你需要什么格式的输出？请选择一项：
  1. 快速聊天简报（5分钟阅读时长，聊天内markdown格式）
  2. 独立文档（带引用的.docx格式，可分享）
提问原因：文档模式会触发更深入的搜索预算和完整审计日志；聊天模式则优先优化交付速度。
强制二选一。

Q3 (asked only if classification ambiguous — ≤1 signal) — Domain disambiguation

Q3（仅当分类模糊时询问——匹配信号≤1）——领域澄清

Quick clarification — pick the closest match:
  1. Academic literature (papers, peer-reviewed)
  2. Industry / trends (what's the buzz, news, sentiment)
  3. Specific entity (a company, person, organization)
  4. Technology / patents (prior art, IP landscape)
  5. Grant funding (NIH, foundations)
  6. Course material (syllabus or curriculum)
  7. None of the above — run general research
Why I'm asking: I couldn't classify confidently from your question alone. This routes you to the right specialist or confirms general-research fallback.
Skip if Q1 + Q2 produced clear specialist match (≥2 signals).
快速澄清，请选择最接近的选项：
  1. 学术文献（论文、同行评审）
  2. 行业/趋势（热点讨论、新闻、舆情）
  3. 特定实体（公司、个人、组织）
  4. 技术/专利（现有技术、IP格局）
  5. 资助资金（NIH、基金会）
  6. 课程资料（教学大纲或课程体系）
  7. 以上都不是——运行通用研究
提问原因：仅通过你的问题无法自信分类，该问题将帮你路由至合适的专业技能，或确认使用通用研究后备工作流。
若Q1+Q2已产生明确的专业技能匹配（信号≥2），则跳过此问题。

Q4 (asked only if Q3 was needed AND user picked "none of the above") — General-research scope

Q4（仅当Q3已询问且用户选择「以上都不是」时提问）——通用研究范围

For general research, what's your time horizon — quick scan (5 searches) or thorough (15 searches)?
Why I'm asking: General research has no specialist budget; you pick it. Quick is good for "what's the lay of the land". Thorough is for "I'll make a decision based on this".
Skip if a specialist took over.
Stop condition: After Q4 (or earlier if dependency skips applied), commit and start Phase 2. Most invocations exit intake after Q1 + Q2.
对于通用研究，你的时间范围是——快速扫描（5次搜索）还是深度调研（15次搜索）？
提问原因：通用研究无专业技能预算，由你选择范围。快速扫描适合了解整体情况；深度调研适合需要据此做决策的场景。
若已委托给专业技能，则跳过此问题。
停止条件：完成Q4后（或因依赖关系提前结束），进入阶段2。大多数调用在完成Q1+Q2后即可结束导入流程。

Phase 2: Deterministic Classification

阶段2：确定性分类

This is deterministic, not LLM-reasoned — for speed, debuggability, and consistency.
python
SIGNALS = {
  pulse:    ["reddit", "hn", "hacker news", "x.com", "twitter", "buzz",
             "sentiment", "trending", "what are people saying",
             "what's happening", "the conversation around",
             "pulse on", "take the pulse", "current conversation"],
  grants:   ["nih", "grant", "grants for", "r01", "r21", "k-award", "reporter",
             "nosi", "funding", "fda", "study section", "principal investigator"],
  litreview:["literature review", "lit review", "litreview", "pico", "spider",
             "systematic review", "review papers on", "research papers on",
             "papers about", "meta-analysis"],
  syllabus: ["syllabus", "course outline", "curriculum", "reading list",
             "for my class", "for my students", "course material"],
  patent:   ["prior art", "fto", "freedom to operate", "patent",
             "patent landscape", "invention", "novelty search",
             "patent search", "ip landscape"],
  dossier:  ["dossier on", "due diligence", "background check",
             "prep me for", "competitor research", "investor diligence",
             "interview prep", "research my competitor", "background on"]
}
此流程为确定性，而非LLM推理——为了速度、可调试性和一致性。
python
SIGNALS = {
  pulse:    ["reddit", "hn", "hacker news", "x.com", "twitter", "buzz",
             "sentiment", "trending", "what are people saying",
             "what's happening", "the conversation around",
             "pulse on", "take the pulse", "current conversation"],
  grants:   ["nih", "grant", "grants for", "r01", "r21", "k-award", "reporter",
             "nosi", "funding", "fda", "study section", "principal investigator"],
  litreview:["literature review", "lit review", "litreview", "pico", "spider",
             "systematic review", "review papers on", "research papers on",
             "papers about", "meta-analysis"],
  syllabus: ["syllabus", "course outline", "curriculum", "reading list",
             "for my class", "for my students", "course material"],
  patent:   ["prior art", "fto", "freedom to operate", "patent",
             "patent landscape", "invention", "novelty search",
             "patent search", "ip landscape"],
  dossier:  ["dossier on", "due diligence", "background check",
             "prep me for", "competitor research", "investor diligence",
             "interview prep", "research my competitor", "background on"]
}

Signals are case-insensitive literal phrases (multi-word substring match).

Signals are case-insensitive literal phrases (multi-word substring match).

Bracketed placeholders (e.g., "research [company]") are intentionally NOT

Bracketed placeholders (e.g., "research [company]") are intentionally NOT

signals — they over-trigger on generic "research X" queries that should

signals — they over-trigger on generic "research X" queries that should

fall back to general research, not auto-route to dossier. Specific phrases

fall back to general research, not auto-route to dossier. Specific phrases

pair the verb with the noun ("dossier on", "background on") and route reliably.

pair the verb with the noun ("dossier on", "background on") and route reliably.

For each specialist S: score[S] = count of SIGNALS[S] phrases matched in question (case-insensitive substring)
if max(score) >= 2: route_to = argmax(score) # high confidence elif max(score) == 1 and only one specialist has score 1: route_to = that specialist # weak match, single specialist else: route_to = "fallback" # ambiguous or no match — ask Q3

**Implementation:** `scripts/classifier.py --question "..."` returns the routing decision + matched signals + per-specialist scores. Use it; don't re-implement.
For each specialist S: score[S] = count of SIGNALS[S] phrases matched in question (case-insensitive substring)
if max(score) >= 2: route_to = argmax(score) # high confidence elif max(score) == 1 and only one specialist has score 1: route_to = that specialist # weak match, single specialist else: route_to = "fallback" # ambiguous or no match — ask Q3

**实现方式**：`scripts/classifier.py --question "..."`返回路由决策、匹配信号及各专业技能得分。请直接使用该脚本，勿重新实现。

Phase 3a: Specialist Delegation (≥2 signals OR single weak match)

阶段3a：专业技能委托（信号≥2 或 单一弱匹配）

When delegating:
  1. Pass the user's question verbatim plus the output preference (Q2)
  2. Let the specialist run its own grill-me intake — do NOT pre-answer specialist questions
  3. Return specialist output as the user-visible result
  4. Tag the result with 
    [Delegated to: research → {specialist}]
    in the chat output so the user knows what skill produced it
  5. Tag the audit log via 
    scripts/routing_transparency_logger.py --action record_delegation
委托时需：
  1. 原样传递用户的问题及输出偏好（Q2）
  2. 让专业技能运行自身的精简式导入流程——勿提前回答专业技能的问题
  3. 将专业技能的输出作为用户可见结果返回
  4. 在聊天输出中标记
    [Delegated to: research → {specialist}]
    ，让用户知晓是哪个技能生成的结果
  5. 通过
    scripts/routing_transparency_logger.py --action record_delegation
    记录审计日志

Phase 3b: Own Fallback Workflow

阶段3b：自有后备工作流

If routing produced no specialist match, run the 8-step fallback.
若路由未匹配到专业技能，则运行8步后备流程。

Step 1: Decompose

步骤1：分解问题

Break the research question into 3–5 sub-questions. Use the framework: what / why / how / who / what's next. Show the decomposition to the user before searching. Use 
scripts/fallback_decomposer.py --question "..."
for a deterministic starting point.
将研究问题拆解为3–5个子问题。使用框架：是什么/为什么/如何实现/谁参与/下一步方向。搜索前需向用户展示分解结果。可使用
scripts/fallback_decomposer.py --question "..."
获取确定性的初始分解方案。

Step 2: Source Selection

步骤2：来源选择

For each sub-question, choose source(s) deterministically:
  • Recency-sensitive → WebSearch + WebFetch + (optionally Reddit/HN if signal)
  • Technical specs / docs → WebSearch + WebFetch
  • Academic → Consensus MCP if connected; otherwise WebSearch with 
    scholar.google.com
    site filter
  • Data / numbers → WebSearch for sources; then WebFetch for primary documents
  • Person / company entity-level → consider routing to 
    dossier
    (offer override)
针对每个子问题，确定性选择来源：
  • 对时效性敏感 → WebSearch + WebFetch +（可选，若有信号则添加Reddit/HN）
  • 技术规格/文档 → WebSearch + WebFetch
  • 学术内容 → 若已连接Consensus MCP则使用；否则在WebSearch中添加
    scholar.google.com
    站点过滤
  • 数据/数字 → WebSearch查找来源；再通过WebFetch获取原始文档
  • 个人/公司实体层面 → 考虑路由至
    dossier
    （提供覆盖选项）

Step 3: Search

步骤3：搜索

Sequential per sub-question. 1 q/sec etiquette. Per source: 2–4 queries, broad-to-narrow.
按子问题顺序执行搜索，遵守每秒1次查询的规范。每个来源执行2–4次查询，从宽泛到精准。

Step 4: Read + Extract

步骤4：读取+提取

For each result that looks high-signal: WebFetch and extract the relevant section. Note the source URL.
对于每个高信号结果：使用WebFetch获取内容并提取相关部分，记录来源URL。

Step 5: Synthesize

步骤5：合成内容

Per sub-question: 2–4 paragraphs answering it with inline citations. Surface disagreement when sources disagree.
针对每个子问题：生成2–4段带内联引用的回答。当来源存在分歧时，需明确展示不同观点。

Step 6: Cross-Cutting Patterns

步骤6：跨子问题模式总结

After per-sub-question synthesis: 1–2 paragraphs of patterns across sub-questions — consensus, controversy, gaps.
完成子问题合成后：生成1–2段内容，总结子问题间的共性模式——共识、争议、研究空白。

Step 7: Output

步骤7：输出

Markdown brief by default (Q2 choice). DOCX if user picked document mode.
默认输出markdown简报（根据Q2选择）；若用户选择文档模式，则输出DOCX格式。

Step 8: Audit Log

步骤8：审计日志

Three-count summary (sent / received / cited) + per-source list with reliability tier (primary / secondary / tertiary).
生成三项计数总结（已发送/已接收/已引用）+ 带可靠性层级（一级/二级/三级）的来源列表。

Routing Transparency Protocol (Mandatory)

路由透明度协议（强制要求）

After classification, the skill always:
  1. States the decision in one sentence: "Routing to 
    litreview
    because you mentioned PICO and meta-analysis (2 signals)."
  2. Offers override: "If you want general research instead OR a different specialist, say so now. Otherwise proceeding in 5 seconds."
  3. Waits 1 turn for confirmation (or auto-proceeds after 5s in interactive contexts).
  4. If user overrides → accept, re-route, log the override via 
    routing_transparency_logger.py --action record_override
    .
Never delegates silently. This is the trust-building property that makes the hybrid pattern work.
分类完成后，该技能必须：
  1. 用一句话说明决策：「因你提到了PICO和meta-analysis（2个信号），将路由至
    litreview
    。」
  2. 提供覆盖选项：「若你想改为通用研究或其他专业技能，请立即告知。否则5秒后继续执行。」
  3. 等待一轮确认（在交互式场景中，5秒后自动继续）。
  4. 若用户覆盖路由 → 接受请求，重新路由，通过
    routing_transparency_logger.py --action record_override
    记录覆盖操作。
绝不静默委托。这是混合模式建立信任的核心特性。

Output Format

输出格式

Markdown brief (Q2 = quick chat briefing)

Markdown简报（Q2 = 快速聊天简报）

markdown
undefined
markdown
undefined

[Research Question] — Briefing

[研究问题] — 简报

Generated: [DATE] | Routed: [delegated specialist | fallback]
生成时间: [DATE] | 路由方式: [委托专业技能 | 后备工作流]

TL;DR

TL;DR

[2-3 sentences]
[2-3句话总结]

Findings

研究发现

[Sub-question 1]

[子问题1]

[2-4 paragraphs with inline citations]
[2-4段带内联引用的内容]

[Sub-question 2]

[子问题2]

...
...

Cross-Cutting Patterns

跨子问题模式

[1-2 paragraphs]
[1-2段内容]

Sources

来源

[Numbered list with hyperlinks, reliability tier per source]
[带超链接的编号列表，标注每个来源的可靠性层级]

Audit

审计信息

[Three counts + per-source tier + failures]
undefined
[三项计数 + 来源层级 + 失败记录]
undefined

DOCX (Q2 = standalone document)

DOCX（Q2 = 独立文档）

Use the standard research-pack DOCX patterns: Arial 12pt, navy headings, blue table headers, hyperlinked sources, mandatory audit log section. Reference the 
docx
skill for setup.
遵循标准研究包DOCX格式：Arial 12号字体，深蓝色标题，蓝色表格表头，超链接来源，强制包含审计日志章节。可参考
docx
技能进行设置。

Audit Log Requirement (Fallback Mode)

审计日志要求（后备模式）

Queries sent:        N
Sources received:    M
Sources cited:       K
Failures:            F (3-consecutive-failures triggered: yes/no)
Per-source tier:     [URL — primary | secondary | tertiary]
Routing decision:    fallback (no specialist matched)
Sub-questions:       [list]
All routing decisions + overrides also logged to 
~/.research_sessions/<session>.json
via 
routing_transparency_logger.py
.
已发送查询数:        N
已接收来源数:    M
已引用来源数:       K
失败次数:            F（是否触发连续3次失败: 是/否）
来源层级:     [URL — 一级 | 二级 | 三级]
路由决策:    后备工作流（未匹配到专业技能）
子问题:       [列表]
所有路由决策+覆盖操作也会通过
routing_transparency_logger.py
记录至
~/.research_sessions/<session>.json
。

Failure Modes

故障模式

FailureBehavior
Classification ambiguous (≤1 signal)Ask Q3 (domain disambiguation).
Specialist delegation failsNote in chat. Offer to retry or fall back to general research.
User overrides routingAccept. Re-route to chosen specialist or fallback. Log the override.
Fallback search returns thin resultsSurface explicitly. Suggest the question may be too niche or too new. Do not fabricate.
3 consecutive tool failures in fallbackStop, alert user, share what was collected.
Question is non-research (e.g., "write me code")Decline politely. Suggest the user invoke an appropriate skill.
Sub-question can't be answeredNote in synthesis as "limited public signal on this"; don't omit silently.
Output format mismatchHonor Q2 preference; if format unavailable, fall back to markdown with note.
Specialist skill missing from environmentSkip it in classification scoring; route to fallback or next-best specialist.
故障类型处理行为
分类模糊（信号≤1）询问Q3（领域澄清）。
专业技能委托失败在聊天中说明，提供重试或切换至通用研究后备工作流的选项。
用户覆盖路由接受请求，重新路由至所选专业技能或后备工作流，记录覆盖操作。
后备搜索结果单薄明确告知用户，提示问题可能过于小众或新颖，不得编造内容。
后备工作流中连续3次工具调用失败停止执行，提醒用户，分享已收集的内容。
非研究类问题（如「帮我写代码」）礼貌拒绝，建议用户调用合适的技能。
子问题无法回答在合成内容中注明「该问题公开信号有限」；不得静默省略。
输出格式不匹配遵循Q2偏好；若格式不可用，则 fallback 至markdown格式并说明。
环境中缺失专业技能在分类评分中跳过该技能，路由至后备工作流或次优专业技能。

Anti-Patterns Rejected

禁用的反模式

  • LLM-reasoned classification (must be deterministic keyword + intent matching)
  • Silent delegation (always surface routing decision)
  • Refusing to route to a specialist when ≥2 signals match
  • Routing to a specialist when classification is genuinely ambiguous (≤1 signal across all)
  • Pre-answering the specialist's grill-me intake (let it run its own)
  • Running fallback when a specialist would clearly do better
  • Fabricating sources in fallback when search is thin
  • Skipping audit log in fallback mode
  • Treating "dossier on [company]" as fallback when 
    dossier
    is the right specialist (the verb-noun-paired phrase, not the generic "research X" form, is what routes)
  • Treating "what are people saying about X" as fallback when 
    pulse
    is the right specialist
  • Auto-routing generic "research [topic]" queries to a specialist when the user hasn't paired the verb with a specialist-specific noun (e.g., "research Microsoft" alone is ambiguous — could be dossier or general; ask Q3 instead of guessing)
  • LLM推理式分类（必须使用确定性关键词+意图匹配）
  • 静默委托（始终展示路由决策）
  • 当信号≥2时拒绝路由至专业技能
  • 当分类确实模糊时（所有技能信号≤1）路由至专业技能
  • 提前回答专业技能的精简式导入问题（让专业技能自行处理）
  • 当专业技能明显更优时运行后备工作流
  • 搜索结果单薄时编造来源
  • 后备模式下跳过审计日志
  • 将「dossier on [company]」视为后备工作流（正确做法是路由至
    dossier
    ，需匹配动词-名词组合短语，而非通用的「research X」形式）
  • 将「what are people saying about X」视为后备工作流（正确做法是路由至
    pulse
    ）
  • 将通用「research [主题]」查询自动路由至专业技能（用户未将动词与专业技能特定名词配对时，如「research Microsoft」本身模糊——可能是dossier或通用研究，应询问Q3而非猜测）

Tooling

工具集

Python (stdlib only)

Python（仅标准库）

  • scripts/classifier.py
     — Deterministic SIGNALS matching → routing decision + per-specialist score + matched phrases. 
    --question "..." --output json
    .
  • scripts/routing_transparency_logger.py
     — JSON-backed audit log at 
    ~/.research_sessions/<session>.json
    . Records every routing decision, override, and delegation handoff.
  • scripts/fallback_decomposer.py
     — Heuristic question → 3–5 sub-questions using what / why / how / who / what's next framework.
  • scripts/classifier.py
     — 确定性SIGNALS匹配 → 路由决策 + 各专业技能得分 + 匹配短语。使用方式：
    --question "..." --output json
    。
  • scripts/routing_transparency_logger.py
     — 基于JSON的审计日志，存储于
    ~/.research_sessions/<session>.json
    。记录所有路由决策、覆盖操作和委托交接。
  • scripts/fallback_decomposer.py
     — 启发式问题分解 → 基于「是什么/为什么/如何实现/谁参与/下一步方向」框架生成3–5个子问题。

Reference Docs (each cites 7+ authoritative sources)

参考文档（每份引用7+权威来源）

  • references/hybrid_router_architecture.md
     — router-vs-run trade-offs + routing transparency principle
  • references/deterministic_classification_canon.md
     — why keyword > LLM-reasoned for routing
  • references/fallback_workflow_canon.md
     — plan-decompose-search-synthesize methodology
  • references/hybrid_router_architecture.md
     — 路由器vs自主执行的权衡 + 路由透明度原则
  • references/deterministic_classification_canon.md
     — 为何路由应使用关键词匹配而非LLM推理
  • references/fallback_workflow_canon.md
     — 规划-分解-搜索-合成方法论

Dependencies

依赖项

  • WebSearch
     + 
    WebFetch
     — Required for fallback workflow
  • Specialist skills — Required for delegation: 
    pulse
    , 
    grants
    , 
    litreview
    , 
    syllabus
    , 
    patent
    , 
    dossier
    . If a specialist is missing, the router skips it in classification and routes to fallback instead.
  • Node.js 
    docx
    library     — Required if user picks document output (Q2 = standalone)
  • Consensus MCP — Optional; used in fallback if academic sub-questions surface
  • WebSearch
     + 
    WebFetch
     — 后备工作流必需
  • 专业技能 — 委托功能必需：
    pulse
    、
    grants
    、
    litreview
    、
    syllabus
    、
    patent
    、
    dossier
    。若某专业技能缺失，路由器会在分类中跳过它，转而路由至后备工作流。
  • Node.js 
    docx
    库     — 用户选择文档输出（Q2 = 独立文档）时必需
  • Consensus MCP — 可选；后备模式下若出现学术子问题则使用

Trigger Phrases

触发短语

  • "research [topic]"
  • "look into [topic]"
  • "what do we know about [topic]"
  • "investigate [topic]"
  • "find me information on [topic]"
  • "do some research on [topic]"
  • "I need to understand [topic]"
  • Any research request that doesn't obviously match a more-specific specialist
Version: 1.0.0 Source spec: 
megaprompts/13-research-megaprompt.md
 Build pattern: Path B (direct conversion)
  • "research [topic]"
  • "look into [topic]"
  • "what do we know about [topic]"
  • "investigate [topic]"
  • "find me information on [topic]"
  • "do some research on [topic]"
  • "I need to understand [topic]"
  • 任何未明显匹配更专业技能的研究请求
版本: 1.0.0 来源规范: 
megaprompts/13-research-megaprompt.md
 构建模式: Path B（直接转换）