deep-research

Deep Research

中文文档

A Claude Code Skill that transforms vague topics into high-quality, deliverable research reports using a systematic 8-step methodology.

The Problem

When researching complex topics, you often face:

• Information overload: Too many sources, hard to distinguish quality
• Fuzzy conclusions: "I feel like X" instead of evidence-based reasoning
• Missing traceability: Can't verify where conclusions came from
• Time-sensitivity blindness: Using outdated information in fast-moving fields like AI

Traditional research is time-consuming, and conclusions are often unreliable.

The Solution

Deep Research provides a systematic 8-step methodology that:

• Tiered source evaluation: L1 (official docs) > L2 (blogs) > L3 (media) > L4 (community)
• Fact cards with citations: Every claim traceable to its source
• Time-sensitivity assessment: Automatic handling of fast-changing fields
• Verifiable conclusions: "Fact → Comparison → Conclusion" explicit chains
• Deliverable output: Boss-readable reports with one-line summaries

Requirements

• Claude Code CLI
• Web search tools (WebSearch, Tavily MCP, or Exa MCP)

Installation

Option 1: Git Clone (Simplest)

git clone https://github.com/wshuyi/deep-research.git
cp -r deep-research/skills/deep-research ~/.claude/skills/

Option 2: Plugin Marketplace

/plugin marketplace add wshuyi/deep-research
/plugin install deep-research@wshuyi/deep-research

Option 3: Third-party Marketplaces

After the repo gets 2+ stars, it will be automatically indexed by SkillsMP. Search for "deep-research" there.

Usage

Natural language triggers:

• "深度调研 [topic]" / "Deep research on [topic]"
• "帮我调研 [topic]" / "Research [topic] for me"
• "对比分析 X 和 Y" / "Compare X and Y"
• "写调研报告" / "Write a research report"

Workflow (8 Steps)

Step 0: Problem type identification
Step 0.5: Time-sensitivity assessment (BLOCKING for AI/tech topics)
Step 1: Problem decomposition & boundary definition
Step 2: Source tiering & authority locking
Step 3: Fact extraction & evidence cards
Step 4: Build comparison framework
Step 5: Reference alignment
Step 6: Fact → Conclusion derivation chain
Step 7: Use case validation (sanity check)
Step 8: Deliverable formatting

Output Structure

~/Downloads/research/<topic>/
├── 00_问题拆解.md          # Problem decomposition
├── 01_资料来源.md          # Source documentation
├── 02_事实卡片.md          # Fact cards
├── 03_对比框架.md          # Comparison framework
├── 04_推导过程.md          # Derivation process
├── 05_验证记录.md          # Validation records
└── FINAL_调研报告.md       # Final deliverable

Example

User: 深度调研 REST API 和 GraphQL 的区别

Claude: [Executes 8-step methodology]
- Identifies as: Concept Comparison type
- Creates fact cards from official specs
- Uses 8-dimension comparison framework
- Validates with real-world scenarios
- Outputs structured report with citations

Key Features

Feature	Description
Source Tiering	L1-L4 hierarchy ensures conclusion traceability
Time-sensitivity	Auto-detects fast-moving fields (AI, crypto, etc.)
Fact Cards	Every claim has source, confidence level, applicability
Explicit Derivation	No "I feel like" - only mechanism-based conclusions
Deliverable Output	One-line summary + structured chapters + citations

FAQ

Q: How is this different from just asking Claude to research something?

A: This skill enforces a systematic methodology with intermediate artifacts, source verification, and explicit reasoning chains. Results are traceable and verifiable.

Q: Does it work for non-technical topics?

A: Yes, the methodology applies to any research topic. The 5 problem types (concept comparison, decision support, trend analysis, problem diagnosis, knowledge organization) cover most use cases.

Q: How does it handle outdated information?

A: Step 0.5 assesses time-sensitivity. For high-sensitivity fields (AI, blockchain), it enforces 6-month time windows and requires version number citations.

License

MIT License - see LICENSE

Author

• wshuyi - GitHub

---

Generated with Claude Code