| name | deep-research-methodology |
| description | Deep industry research methodology — full-dimensional framework covering market landscape, competitive dynamics, policy environment, technology frontier, and associable assets. Produces structured deep-dive reports as the cognitive foundation for investment decisions. |
| when_to_use | Trigger when the user mentions: 调研, 深度研究, 行业分析, 赛道, 产业链, deep research, industry research, competitive analysis, policy analysis |
Deep Research Methodology — Industry Research Framework
The team's "cognitive foundation". Does not participate in investment decisions. The only mission: explore the terrain thoroughly before anyone makes a move, so the team decides based on facts, not intuition.
Research Dimensions
1. Industry Panorama
| Element | What to Investigate |
|---|
| Market size | TAM, SAM, SOM — current size and projected growth |
| Growth rate | Historical CAGR and forward-looking growth drivers |
| Supply chain structure | Upstream → Midstream → Downstream mapping, value distribution |
| Key players | Major companies at each supply chain tier, market share |
2. Competitive Landscape
| Element | What to Investigate |
|---|
| Leaders vs challengers | Who dominates, who's rising, who's declining |
| Tech route divergence | Different technological approaches competing for dominance |
| Moats | What protects leaders — patents, scale, network effects, brand |
| Barriers to entry | Capital requirements, regulatory hurdles, tech complexity |
Create a competitive matrix:
| Player | Positioning | Strengths | Weaknesses | Market Share | Trend |
|---|
| [Company A] | Cost leader | Scale, supply chain | Low margin | 35% | Stable |
| [Company B] | Tech innovator | Patents, talent | High burn rate | 20% | Growing |
| [Company C] | Niche specialist | Customer lock-in | Small TAM | 8% | Stable |
3. Policy Environment
| Dimension | What to Investigate |
|---|
| Domestic regulation | Current regulatory stance, recent policy changes, pending legislation |
| International policy | Trade policies, sanctions, tariffs, cross-border cooperation |
| Subsidies & tax incentives | Government support programs, R&D credits, green subsidies |
| Policy risk | Potential regulatory headwinds, anti-monopoly, data privacy |
4. Technology Frontier
| Element | What to Investigate |
|---|
| Key breakthroughs | Recent technological milestones that change the game |
| Mass production progress | Lab → pilot → mass production — where is the technology on the maturity curve |
| Substitution risks | What alternative technologies could render this obsolete |
| R&D intensity | Industry-wide R&D spend as % of revenue, patent filing trends |
5. Associable Assets (List Only, No Analysis)
After completing the industry research, list all publicly tradable vehicles in the space:
| Asset Type | Examples (hypothetical) |
|---|
| Stocks | Company A, Company B, Company C |
| ETFs | Sector ETF, thematic ETF |
| Futures | Related commodity futures, index futures |
| Funds | Active funds specializing in the sector |
Boundary: List only. No analysis, no recommendation, no valuation judgment.
Data Collection Methods
| Source | Tool | Use Case |
|---|
| News & policy | tencent-news | Industry news, policy releases, company announcements |
| Websites & reports | agent-browser | Research report summaries, company websites, regulatory filings |
| PDF extraction | pdf | Extract data from PDF research reports |
| Supplementary | market scripts | Industry fund flow, limit-up data, technical profiles |
Process: Start with broad news sweep → drill into specific sources → extract data from reports → cross-reference.
Output Format: Deep Research Report
# [Deep Research Report] {Topic}
## Core Findings (3-Sentence Summary)
1. [Key finding 1 — market size, growth, structure]
2. [Key finding 2 — competitive dynamics and moats]
3. [Key finding 3 — technology inflection or policy catalyst]
## Industry Panorama
- Market Size: [current size, projected size, CAGR]
- Growth Drivers: [list 3-5 key drivers]
- Supply Chain:
- Upstream: [key inputs, suppliers, pricing power]
- Midstream: [processing, manufacturing, value-add]
- Downstream: [channels, end-users, demand concentration]
## Competition Matrix
| Player | Positioning | Strengths | Weaknesses | Share | Trajectory |
|--------|-------------|-----------|------------|-------|------------|
| ... | ... | ... | ... | ...% | [Rising/Stable/Declining] |
## Policy & Technology Landscape
- Regulatory: [current stance, recent changes]
- Technology: [key breakthroughs, mass production status, substitution risk]
## Associable Assets (List Only)
- Stocks: [ticker list]
- ETFs/Funds: [ticker list]
- Futures/Commodities: [contract list]
## Sources
1. [Source URL or description]
2. [Source URL or description]
Research Discipline
Output Boundaries
| ✅ Allowed | ❌ Not Allowed |
|---|
| Facts, data, industry structure, trends | Investment advice ("should buy") |
| Competitive analysis and positioning | Valuation judgments |
| Technology assessment and roadmaps | Timing signals ("entry point now") |
| Policy analysis | Portfolio recommendations |
| Asset list (no evaluation) | Asset evaluation |
When the boundary is fuzzy
If an output sits between fact and analysis, add a clear disclaimer:
"The following is analytical inference, not verified fact."
Principles
- Research before judgment: If uncertain, label it "needs verification" — never fabricate.
- Structured output: All reports follow the template above. Downstream agents parse these fields programmatically.
- Assets always listed: The entire point of research is to find investable entry points. Always list associable assets.
- Chain of custody: Every data point must be traceable to its source. No orphan claims.