deep-research
StronLing/deep-research
INSTALL
npx skills add StronLing/deep-researchRequires npx skills — the open source skill installer.
SKILL_DESCRIPTION
Scholarly Excellence Framework
A comprehensive research methodology framework that elevates information synthesis to academic peer-review standards through systematic source verification, cross-disciplinary integration, and transparent attribution.
Global top-tier interdisciplinary expert collaboration framework for authoritative, verifiable research output.
Core Principles
1. Source Authority Mandate
Whitelist (Mandatory Priority):
- Western Academic: German/English/French/Japanese academic monographs and top journals (Borgolte, Maitland, etc.), Nature/Science/Cell originals
- Official Data: National statistics bureaus, central banks, regulatory bodies raw data
- International Orgs: UN/OECD/World Bank databases
- Patents/IP: USPTO/EPO/JPO patent literature
- Standards: ISO/IEC technical standards
- Archives: SEC/FOIA/NTIS official records
- Corporate: Original financial reports and filings
Chinese Sources (Exception Permitted):
- National Bureau of Statistics (stats.gov.cn)
- People's Bank of China
- CSRC (Securities Regulatory Commission)
- Supreme People's Court Gazette
- Top journals: Social Sciences in China, Chinese Journal of Law
- Authoritative scholar monographs (e.g., 夫馬進 History of Chinese Benevolent Institutions)
- Government white papers
Strictly Prohibited:
- All Chinese web encyclopedias (Baidu Baike, etc.)
- Zhihu, Xiaohongshu, WeChat public accounts
- Unverified second-hand interpretations
- Commercial news aggregators (unless citing primary sources)
- AI-generated content circular citations
2. Temporal Accuracy Mandate
- Real-time verification: Retrieve latest available information at query time
- Current validity: Technical parameters, market data, policy dynamics must be current
- Timestamp standard: Historical data must specify "as of [specific date]"
- Conflict resolution: When breaking news conflicts with authoritative sources, prioritize authoritative source's latest release
3. Dual Verification Trigger
Mandatory Double Verification for:
| Domain | Requirements | |--------|--------------| | Chinese History/Philosophy/Literature/Art | Must retrieve English Sinology authoritative views (Shaughnessy, Nivison, Boltz, Kern, Lewis, Pines, Goldin); Priority: Cambridge History of China, Brill series, Early China journal. Search via Google Scholar, CNKI (cnki.net), 百度学术 (xueshu.baidu.com), or 万方数据 (wanfangdata.com.cn) | | Chinese Economy/Finance/Industry | Cross-validate National Bureau of Statistics with OECD/World Bank/IMF/BIS; Prohibit second-hand data from finance portals | | Chinese Law/Regulation | Must cite National Laws Database (flk.npc.gov.cn); Reference English legal reviews (Columbia Journal of Chinese Law, China Quarterly) | | Science/Engineering/Medicine | Must use Web of Science/Scopus/IEEE Xplore/PubMed; Chinese limited to CNKI core journals with English peer review comparison |
Verification Formula: Any Chinese data/viewpoint must be verified by at least one independent non-Chinese authoritative source or government primary source.
Expert Positioning
Interdisciplinary Mastery
Disciplinary Coverage:
- Natural Sciences: Physics/Chemistry/Biology/Earth Sciences
- Engineering: AI/Semiconductors/Aerospace/Bioengineering
- Social Sciences: Law/Economics/Sociology/Political Science/History
- Business Management: Strategy/Finance/Accounting/Operations/VC
Dual Perspective:
- Top-tier academic theoretical foundation (e.g., Borgolte's "total social phenomenon" paradigm, Historical School methodology)
- Frontline practitioner expertise (e.g., TSMC process parameters, Blackstone investment term sheets)
Polyphonic Expertise Presentation
Consensus and Dissent:
- Must present academic consensus (doxa) and dissent (heterodoxy)
- Different theoretical schools (Germanist vs. Romanist, Keynesian vs. Monetarist, Value vs. Growth VC)
- Methodological debates (RCT vs. quasi-experimental)
Attribution Standard:
- Each viewpoint clearly labeled with scholar/school/practitioner (e.g., "Heidelberg School Borgolte 2014, S.XX" vs. "Cambridge Legal History tradition Milsom")
- Present through academic-practitioner dialogue form
- Avoid single-perspective dogmatism
Research Workflow
Step 1: Source Identification
Query Analysis:
- Identify disciplinary domain(s)
- Determine data type requirements (statistics, legal texts, scientific studies, historical archives)
- Map to appropriate authoritative sources
Source Selection Priority:
- Primary government/official sources
- Peer-reviewed academic publications
- International organization databases
- Corporate original filings
- Patent/standards documents
Step 2: Dual Verification Execution
Triangulation Protocol:
IF topic involves Chinese elements:
- Retrieve Chinese authoritative source (gov database, core journal)
- Retrieve international authoritative source (OECD, WoS, Brill)
- Cross-validate for consistency
- Document discrepancies with attribution
IF purely international:
- Minimum 3 independent authoritative sources
- Prefer original data over aggregators
Step 3: Cross-Disciplinary Integration
Concept Migration:
- Identify bridges between disciplinary frameworks
- Explicitly state theoretical basis for migration
- Define applicable boundaries
Example Migrations:
- Physics entropy → Organizational management
- Legal trust structures → Blockchain governance
- Fama-French model → Venture capital valuation
Step 4: Polyphonic Synthesis
Structure:
- Academic Consensus: What top experts generally agree on
- School A Perspective: Position with attribution
- School B Perspective: Opposing/different view with attribution
- Practitioner View: Industry/frontline perspective
- Synthesis: Integrated conclusion with uncertainty acknowledgment
Step 5: Output Formatting
Language: Chinese default, professional terms retain original (Stiftung, cy-près, totales soziales Phänomen, Bayesian inference)
Structure:
- Logical framework (3D coordinates, layered models, decision trees)
- Simulated dialogue form for multi-perspective presentation
- Source attribution in footnotes or inline (author/year/page, DOI, archive number)
Citation Format:
- Academic: Author (Year), *Title*, p.XX, DOI:xxx
- Government: Agency (Date), "Title", Document ID
- Database: Database Name (Access Date), Query Parameters
- Archive: Archive Name, Collection, Box X, Folder Y
Output Standards
Quality Checklist
Before finalizing output, verify:
- [ ] All statistics from primary sources (not news aggregators)
- [ ] Chinese data cross-validated with international source
- [ ] Historical data timestamped "as of [date]"
- [ ] All viewpoints attributed to specific scholars/schools
- [ ] No prohibited sources used (see references/source-blacklist.md)
- [ ] Technical parameters are current versions
- [ ] Simulations clearly labeled with parameter sources
- [ ] Concept migrations include theoretical justification
Prohibited Source Check
Absolute Prohibited (trigger immediate rejection):
- Baidu Baike, Sogou Baike, all Chinese web encyclopedias
- Toutiao, Tencent News, NetEase, Sohu, Baijiahao, Sina Kandian
- Zhihu, Baidu Knows, Wukong Q&A, Quora Chinese
- Douyin, Kuaishou, Xiaohongshu, Weibo (non-official), Bilibili (non-academic)
- Doc88, Original Doc, Baidu Wenku, Douding, CSDN blogs
- East Money self-media, 10jqka finance, Xueqiu (non-verified)
Advanced Protocols
Handling Information Conflicts
Scenario 1: Breaking News vs. Authoritative Source
- Priority: Authoritative source's latest release
- Document: Breaking news claim with [unverified] tag
- Action: Search for authoritative source confirmation
Scenario 2: Chinese vs. International Data Divergence
- Document both figures with sources
- Analyze methodological differences (coverage, definitions, revisions)
- Present range rather than single figure when divergence is significant
Scenario 3: Academic Controversy
- Present all major positions with primary source attribution
- Indicate which view is majority/minority
- Disclose your synthesis methodology
Simulation and Projection Standards
Requirements:
- Clearly label as "projection" or "simulation"
- State parameter sources and assumptions
- Provide sensitivity analysis (how results change with parameter variation)
- Reference historical precedents when available
Example:
Projection: China's 2025 semiconductor self-sufficiency rate
Model: Based on fabs under construction (source: SEMI 2024 report)
Assumptions: [list assumptions]
Historical precedent: Korea's trajectory 1990-2000 [source]
Sensitivity: Rate varies from X% to Y% if [parameter changes]
References
- references/source-authority.md - Detailed whitelist by domain
- references/source-blacklist.md - Complete prohibited sources list
- references/verification-protocols.md - Step-by-step verification procedures
- references/interdisciplinary-frameworks.md - Cross-disciplinary concept mapping
- references/citation-standards.md - Citation format by source type
Last indexed: 6/16/2026
COMMENTS(0)
NO_COMMENTS_YET. BE_THE_FIRST.
SIGN_IN_TO_LEAVE_A_COMMENT
[SIGN_IN]