Skill Positioning
This skill is used togenerate a structured research outline before institutional research on a listed company, helping researchers systematically prepare research questions, covering key dimensions such as finance, business, strategy, and risk.with gildata-aidata financial data services as the core data source, ensuring the professionalism and accuracy of financial data, valuation data, and peer comparisons.
Core Advantages：
- ✅ gildata-aidata professional data source: all financial data, valuation data, and consensus estimates are preferentially obtained through gildata-aidata services
- ✅ Systematic analysis framework: covers four major modules: fundamentals, peer comparison, historical issues, and core issues
- ✅ Professional question design: 10-12 targeted research questions, avoiding repetition of public information
- ✅ Structured output: standardized Markdown-format report, directly displaying the complete content in the conversation
Output Format：
- By default, generate a structured Markdown-format report
- The report content is directly displayed in the conversation, with no need to send a file
- If the user explicitly requests a Word document, it can be handled separately
Trigger Scenarios
This skill should be triggered when any of the following expressions appears from the user:
| User expression | Trigger type |
|---|---|
| "Generate a research outline for Company XX" | Targeted generation |
| "Prepare to visit Company XX for research" | Targeted generation |
| "What questions need to be asked before researching Company XX" | Targeted generation |
| "Help me write a research outline" | Targeted generation |
| "Institutional research preparation materials for Company XX" | Targeted generation |
| "Question list for first-time research on Company XX" | Targeted generation |
| "Prepare to research Company XX" | Targeted generation |
Note: The user only needs to mention the company name + keywords such as research/outline/outline, and this skill should be triggered.
Execution Flow
Step 1: Obtain basic company information
1.1 Extract company name/stock code
Identify the target listed company from the user input:
- Company full name/abbreviation (such as "Wuliangye", "Moutai")
- Stock code (such as "000858", "600519")
1.2 Obtain financial and valuation data through gildata-aidata
Core principle: all financial data, valuation data, and consensus estimates are preferentially obtained through gildata-aidata services.
| Data category | gildata-aidata tool | Example query parameters |
|---|---|---|
| Financial statements | FinancialStatement |
"Wuliangye 2024 annual report income statement balance sheet cash flow statement" |
| Financial analysis indicators | FinancialAnalysis |
"Wuliangye 2024 gross margin, net margin, ROE, ROIC, debt-to-asset ratio, accounts receivable turnover ratio, inventory turnover ratio" |
| Main business composition | MainOperIncData |
"Wuliangye 2024 by product, by industry, by region, revenue, cost, gross margin" |
| Consensus estimates | ConsensusExpectation |
"Wuliangye 2024 2025 2026 revenue net profit EPS consensus estimates" |
| Valuation data | StockValueAnalysis |
"Wuliangye PE PB PS EV dividend yield" |
| Institutional ratings | InstitutionalRating |
"Wuliangye institutional ratings rating score" |
| Earnings report commentary reference | AShareFinancialReportReview |
"Wuliangye 2024 annual report performance commentary" |
Hong Kong stock companies: use HKFinancialStatement、HKStockValueAnalysis、HKMainOperIncData
US stock companies: use USFinancialStatement、USValuationMetrics、USFinancialAnalysis
1.3 Obtain supplementary information via web_search
Used for information that gildata-aidata cannot cover:
- Basic company information (founding time, employee size, headquarters location, etc.)
- Historical research records/investor relations activity records
- Latest news sentiment, industry trends
web_search --query "{公司名称} 公司简介 主营业务 行业地位" --numResults 10
web_search --query "{公司名称} 投资者关系活动记录 机构调研 问答 2025 2026" --numResults 10
Step 2: Fundamental analysis
2.1 Company overview
Organize the following information:
- Company full name, stock code, listing exchange
- Founding time, listing time
- Main business and products
- Industry classification and sub-sector
- Market capitalization, employee size
2.2 Financial analysis (obtained via gildata-aidata)
Analyze the following core indicators (most recent 3 years + latest period):
| Indicator category | Specific indicators | Data source |
|---|---|---|
| Growth | Revenue growth rate, net profit growth rate, non-recurring net profit growth rate | FinancialStatement |
| Profitability | Gross margin, net margin, ROE, ROIC | FinancialAnalysis |
| Operating capability | Accounts receivable turnover ratio, inventory turnover ratio, total asset turnover ratio | FinancialAnalysis |
| Solvency | Debt-to-asset ratio, current ratio, quick ratio, interest coverage ratio | FinancialAnalysis |
| Cash flow | Net operating cash flow, free cash flow, cash flow/net profit ratio | FinancialStatement |
2.3 Business analysis (obtained via gildata-aidata)
- Main business revenue composition (by product/region/channel) →
MainOperIncData - Core products/services and competitive advantages
- Concentration of major customers and suppliers
- Production capacity and utilization
- R&D investment and core technologies
2.4 Industry position
- Market share and ranking
- Main competitors
- Industry concentration (CR3/CR5)
- The company's position in the industry chain (upstream/midstream/downstream)
Step 3: Peer comparison analysis
3.1 Determine comparable companies
Select 3-5 listed companies in the same industry as comparison targets, with selection criteria:
- Same sub-industry
- Similar business model
- Comparable market capitalization (or industry leader)
- A-shares/Hong Kong stocks/US stocks preferred
3.2 Obtain comparison data via gildata-aidata
Call the following tools for the target company and comparable companies respectively:
| Comparison dimension | Specific indicators | gildata-aidata Tool |
|---|---|---|
| Valuation Level | PE(TTM)、PB、PS、PEG、EV/EBITDA | StockValueAnalysis |
| Growth | 3-year revenue CAGR, 3-year net profit CAGR | FinancialStatement |
| Profitability | Gross margin, net margin, ROE | FinancialAnalysis |
| Operating Efficiency | Inventory turnover days, accounts receivable turnover days | FinancialAnalysis |
| Financial Health | Debt-to-asset ratio, operating cash flow/revenue | FinancialAnalysis |
3.3 Comparison Conclusion
- The company's valuation level within the industry (high/medium/low)
- Growth ranking
- Profitability ranking
- Summary of competitive advantages and disadvantages
Step 4: Review of Historical Research Questions
4.1 Search Historical Research Records
Use web_search to query recent investor relations activity records:
web_search --query "{公司名称} 投资者关系活动记录表 机构调研 问答" --numResults 10
4.2 Organize Historical Questions
Extract from the most recent investor relations activity records (3-5 times):
- High-frequency questions of institutional concern
- Key points of management responses
- Unresolved questions not clearly answered
- Main directions covered by the questions (strategy/finance/business/risk)
4.3 Question Classification
Classify historical questions into the following categories:
- Strategic planning
- Business development
- Financial condition
- Risk factors
- Industry competition
Step 5: Generate Research Question List
5.1 Question Design Principles
| Principle | Description |
|---|---|
| Targeted | Combine with the company's specific situation, avoid generic questions |
| Hierarchical | From strategy to execution, from macro to micro |
| Open-ended | Use more "how" and "why" to guide in-depth answers |
| Verifiable | Answers to questions can be verified through subsequent data |
| Differentiated | Avoid repeating public information, focus on expectation gaps |
5.2 Number of Questions
Generate10-12core research questions, covering the following dimensions:
| Dimension | Number of Questions |
|---|---|
| Strategic planning | 2-3 |
| Business development | 3-4 |
| Financial condition | 2-3 |
| Risk factors | 1-2 |
| Industry competition | 1-2 |
5.3 Question Quality Requirements
Each question should include:
- Question Background: Briefly explain the reason/basis for asking
- Core Question: Clear and specific question statement
- Follow-up Direction: Possible follow-up points
Step 6: Generate Research Outline Report
6.1 Organize Content Structure
Organize the report content according to the [Report Structure Template] below.
6.2 Format Output
- Use Markdown heading levels (#, ##, ###) to organize sections
- Use tables to display financial data and comparison information
- Use lists to present key points and risk alerts
- Ensure clear logic and distinct hierarchy
6.3 Output Report
Display the complete structured report directly in the conversation, no need to send a file.
Report Structure Template
# 📋 {公司名称} 机构调研大纲
**调研类型**：首次调研
**准备时间**：标准版（1-2 小时）
**生成日期**：YYYY 年 MM 月 DD 日
---
## 一、公司基本信息
| 项目 | 内容 |
|------|------|
| 公司全称 | [全称] |
| 股票代码 | [代码].SH/.SZ/.HK |
| 上市交易所 | [交易所] |
| 成立时间 | [年份] |
| 上市时间 | [年份] |
| 法定代表人 | [姓名] |
| 总部地点 | [城市] |
| 员工规模 | [人数] |
| 总市值 | [亿元] |
| 所属申万行业 | [行业分类] |
### 主营业务
[描述公司主营业务、核心产品/服务、应用场景]
### 行业地位
[描述公司在行业中的市场地位、市场份额、排名等]
---
## 二、基本面分析
### 2.1 财务概况
| 财务指标 | 2023 年 | 2024 年 | 2025 年 | 2026Q1 |
|----------|--------|--------|--------|--------|
| 营业收入（亿元） | | | | |
| 营收增速（%） | | | | |
| 归母净利润（亿元） | | | | |
| 净利润增速（%） | | | | |
| 毛利率（%） | | | | |
| 净利率（%） | | | | |
| ROE（%） | | | | |
| 资产负债率（%） | | | | |
| 经营性现金流（亿元） | | | | |
### 2.2 财务分析要点
**成长性分析**：
[分析营收和净利润增长趋势，识别增长驱动因素]
**盈利能力分析**：
[分析毛利率、净利率、ROE 变化趋势及原因]
**营运能力分析**：
[分析应收账款、存货周转情况]
**偿债能力分析**：
[分析资产负债率、流动比率等指标]
**现金流分析**：
[分析经营性现金流与净利润匹配度]
### 2.3 业务结构
| 业务板块 | 营收占比 | 毛利率 | 增速 |
|----------|----------|--------|------|
| [业务 1] | % | % | % |
| [业务 2] | % | % | % |
| [业务 3] | % | % | % |
### 2.4 核心竞争优势
1. **[优势 1]**：[具体描述]
2. **[优势 2]**：[具体描述]
3. **[优势 3]**：[具体描述]
---
## 三、同业对比分析
### 3.1 可比公司选择
| 公司名称 | 股票代码 | 市值（亿元） | 选择理由 |
|----------|----------|--------------|----------|
| [公司 A] | | | |
| [公司 B] | | | |
| [公司 C] | | | |
| [公司 D] | | | |
### 3.2 估值对比
| 公司 | PE(TTM) | PB | PS | PEG | EV/EBITDA |
|------|---------|-----|-----|-----|-----------|
| {目标公司} | | | | | |
| [公司 A] | | | | | |
| [公司 B] | | | | | |
| [公司 C] | | | | | |
| [公司 D] | | | | | |
| **行业中位数** | | | | | |
### 3.3 成长性对比
| 公司 | 营收 3 年 CAGR | 净利润 3 年 CAGR |
|------|----------------|------------------|
| {目标公司} | % | % |
| [公司 A] | % | % |
| [公司 B] | % | % |
| [公司 C] | % | % |
| [公司 D] | % | % |
### 3.4 盈利能力对比
| 公司 | 毛利率（%） | 净利率（%） | ROE（%） |
|------|-------------|-------------|----------|
| {目标公司} | | | |
| [公司 A] | | | |
| [公司 B] | | | |
| [公司 C] | | | |
| [公司 D] | | | |
### 3.5 对比结论
**估值水平**：[高/中/低] 于同业，原因分析...
**成长性**：[领先/持平/落后] 同业，原因分析...
**盈利能力**：[领先/持平/落后] 同业，原因分析...
**综合评估**：[总结公司在同业中的竞争地位]
---
## 四、历史调研问题回顾
### 4.1 近期调研记录
| 调研日期 | 调研形式 | 参与机构数量 | 来源 |
|----------|----------|--------------|------|
| [日期] | 现场调研/电话会议 | [数量] | [公告/网站] |
| [日期] | 现场调研/电话会议 | [数量] | [公告/网站] |
| [日期] | 现场调研/电话会议 | [数量] | [公告/网站] |
### 4.2 高频关注问题
| 问题方向 | 具体问题 | 管理层回复要点 |
|----------|----------|----------------|
| [方向 1] | [问题] | [要点] |
| [方向 2] | [问题] | [要点] |
| [方向 3] | [问题] | [要点] |
### 4.3 遗留问题/未明确事项
1. [问题 1]：[说明为何未明确]
2. [问题 2]：[说明为何未明确]
3. [问题 3]：[说明为何未明确]
---
## 五、核心调研问题清单（10 问）
### 5.1 战略规划类（2-3 问）
**问题 1**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
**问题 2**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
**问题 3**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
### 5.2 业务发展类（3-4 问）
**问题 4**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
**问题 5**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
**问题 6**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
**问题 7**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
### 5.3 财务状况类（2-3 问）
**问题 8**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
**问题 9**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
**问题 10**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
### 5.4 风险因素类（1-2 问）
**问题 11**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
### 5.5 行业竞争类（1-2 问）
**问题 12**：[问题标题]
- **问题背景**：[背景说明]
- **核心问题**：[具体问题表述]
- **追问方向**：[可能的追问点]
---
## 六、调研准备清单
### 6.1 资料准备
- [ ] 公司年报/季报（最近 3 年 + 最新一期）
- [ ] 招股说明书（如为次新股）
- [ ] 近期投资者关系活动记录表
- [ ] 券商研报（最近 3 个月）
- [ ] 同业公司财报
- [ ] 行业研究报告
- [ ] 公司公告（重大事项）
### 6.2 数据准备
- [ ] 财务数据对比表
- [ ] 估值对比表
- [ ] 股价走势图（1 年/3 年/5 年）
- [ ] 机构持仓变化
- [ ] 北向资金持仓（如为沪深股通标的）
### 6.3 问题准备
- [ ] 核心问题清单（已生成）
- [ ] 追问问题清单
- [ ] 数据验证问题
- [ ] 敏感问题备选
---
## 七、风险提示
1. **信息时效性**：本大纲基于公开信息生成，部分数据可能存在滞后，建议调研前更新最新数据
2. **行业差异**：本大纲为通用模板，特定行业（如银行、保险、地产）需补充行业特有指标
3. **调研礼仪**：避免询问未公开的重大敏感信息，遵守信息披露相关规定
4. **独立判断**：管理层回复可能存在主观倾向，需结合多方信息独立验证
---
## 八、调研后跟进事项
1. **整理调研纪要**：24 小时内完成调研记录整理
2. **更新盈利预测**：根据调研信息调整财务模型
3. **撰写点评报告**：形成书面调研总结
4. **投资决策建议**：给出买入/持有/卖出建议及理由
5. **持续跟踪**：建立跟踪清单，定期验证调研信息
---
> **生成说明**：本大纲由 AI 助手基于公开信息自动生成，仅供参考。实际调研请结合最新信息和专业判断。
> **免责声明**：本大纲不构成投资建议，市场有风险，投资需谨慎。
Special Scenario Handling
Scenario 1: User does not provide a company name
When the user only says "generate a research outline" but does not specify a company:
- Proactively ask the user which listed company they need to research
- Can provide recently popular research companies for reference
Scenario 2: Limited company information (recently listed/small-cap)
When the target company is a recently listed or small-cap company with limited public information:
- Mark "limited information" in the report
- Rely mainly on the prospectus and latest financial reports
- Add an "Information Acquisition Suggestions" section
- Add basic information confirmation questions to the research questions
Scenario 3: Cross-industry company
When the target company's business spans multiple industries:
- Analyze each business segment separately
- Select comparable companies for each business segment
- Research questions cover each business segment
Scenario 4: Hong Kong/US listed company
When the target company is a Hong Kong or US listed company:
- Adjust financial indicator units (RMB/USD/HKD)
- Select comparable companies in the corresponding market
- Note differences in disclosure requirements across markets
- Use the corresponding Hong Kong/US gildata-aidata tools
Tool Usage Instructions
| Tool | Purpose | Call Example |
|---|---|---|
| finx gildata-aidata | Core data source, to obtain financial statements, financial analysis indicators, main business composition, valuation, consensus estimates, etc. | finx gildata-aidata FinancialStatement --body '{"query": "五粮液 2024 年 利润表"}' |
web_search |
Supplementary queries, used for basic company information, historical research records, news and public opinion, etc. | web_search --query "五粮液 投资者关系活动记录" --numResults 10 |
web_fetch |
Obtain webpage details (announcements, research reports) | web_fetch --url "网址" |
Important Principles：
- All financial data, valuation data, and consensus estimate queries must prioritize the gildata-aidata service
- If gildata-aidata data is incomplete, web_search queries may be used as a supplement
- Data source priority: gildata-aidata > company announcements > brokerage research reports > financial news
gildata-aidata Tool Call Guide
Common Query Templates for A-share Companies
# 财务报表（三大表）
finx gildata-aidata FinancialStatement --body '{"query": "{公司名} 2024 年年报 利润表 资产负债表 现金流量表"}'
# 财务分析指标
finx gildata-aidata FinancialAnalysis --body '{"query": "{公司名} 2024 年 毛利率 净利率 ROE ROIC 资产负债率 应收账款周转率 存货周转率"}'
# 主营构成
finx gildata-aidata MainOperIncData --body '{"query": "{公司名} 2024 年 分产品 分行业 分地区 收入 成本 毛利率"}'
# 一致预期
finx gildata-aidata ConsensusExpectation --body '{"query": "{公司名} 2024 2025 2026 营收 净利润 EPS 一致预期"}'
# 估值数据
finx gildata-aidata StockValueAnalysis --body '{"query": "{公司名} PE PB PS EV 股息率"}'
# 机构评级
finx gildata-aidata InstitutionalRating --body '{"query": "{公司名} 机构评级 评级分数"}'
# 财报业绩点评参考
finx gildata-aidata AShareFinancialReportReview --body '{"query": "{公司名} 2024 年年报 业绩点评"}'
Common Query Templates for Hong Kong Companies
# 港股财务报表
finx gildata-aidata HKFinancialStatement --body '{"query": "{公司名} 2024 年 财务报表"}'
# 港股估值
finx gildata-aidata HKStockValueAnalysis --body '{"query": "{公司名} PE PB 股息率"}'
Common Query Templates for US Companies
# 美股财务报表
finx gildata-aidata USFinancialStatement --body '{"query": "{公司名} 2024 年 财务报表"}'
# 美股估值
finx gildata-aidata USValuationMetrics --body '{"query": "{公司名} PE PB PS"}'
Data Source Priority
| Data Type | First Priority | Second Priority | Third Priority |
|---|---|---|---|
| Core financial data | gildata-aidata FinancialStatement |
Company announcements (CNINFO) | web_search |
| Financial analysis indicators | gildata-aidata FinancialAnalysis |
East Money | web_search |
| Business segment data | gildata-aidata MainOperIncData |
Company announcements | Brokerage research reports |
| Consensus estimates | gildata-aidata ConsensusExpectation |
East Money | web_search |
| Valuation data | gildata-aidata StockValueAnalysis |
East Money | Hithink RoyalFlush |
| Institutional ratings | gildata-aidata InstitutionalRating |
Brokerage research reports | web_search |
| Basic company information | web_search | Company announcements | CNINFO |
| Historical Research Records | web_search | Exchange Interactive Easy | Company Announcements |
Quality Checklist
Before outputting the report, check the following items:
- Company basic information is accurate (name, code, industry)
- Financial data is labeled with reporting period and source (prefer gildata-aidata)
- Peer companies are reasonably selected (same sub-industry)
- Comparative data uses consistent calibers (TTM/annual report/quarterly report)
- Research questions are targeted (not generic questions)
- The number of questions is in the range of 10-12
- Historical research questions have source support
- Risk disclosures are sufficient
- The report format complies with the template
- A disclaimer has been added
- The report is displayed directly in the conversation and has not been written to a file
Notes
- Data Accuracy: Financial data must be labeled with source and reporting period, and gildata-aidata should be used preferentially for retrieval
- Objective and Neutral: Analysis should be based on facts and avoid subjective assumptions
- Information Timeliness: Prefer the latest public information (within the last 12 months)
- Compliance Notice: Remind users to comply with relevant research regulations and not obtain insider information
- Question Quality: Research questions should avoid answers that can be obtained from public information
- Peer Selection: Comparable companies should be the 3-5 with the closest business models
- Industry Differences: General templates may not apply to all industries, and users need to be reminded
- Data Sources: Financial/valuation/consensus estimates should preferentially use gildata-aidata; basic information/research records use web_search
- Conciseness First: Keep the report at a moderate length and highlight the key points
- Continuous Updates: Remind users to update the latest data before research
- Output Method：The report content is displayed directly in the conversation, with no need to write to a file or send it to the user
Skill Update Log
2026-05-11 Update：
- Core Data Source Change: Change all financial data, valuation data, and consensus estimate queries to preferentially use the gildata-aidata service
- New Tool Matrix: Establish a gildata-aidata tool query reference table for A-share, Hong Kong stock, and U.S. stock companies
- Improved Invocation Guide: Provide complete instructions for the gildata-aidata tool invocation format and common query templates
- Optimized Data Retrieval Order: Step 1 clarifies that gildata-aidata is prioritized for obtaining financial and valuation data, while web_search is only used for supplementary information
- Updated Data Source Priority: Raised gildata-aidata to the first-priority data source for financial/valuation/consensus estimates
2026-04-08 Update：
- Clarified the report output method: display directly in the conversation, with no need to send a file
- Optimized the report structure template
- Updated the notes and output method description
Skill Version：v2.0
Last Updated：2026-05-11
Maintainer: Dianjin Intelligent Assistant