Install
openclaw skills install company-investment-researchStructured, multi-dimensional company investment research framework for AI agents and human analysts. Provides a 10-part checklist (moat, tech, market, customers, growth, financials, geography, governance, valuation, recommendation) to turn scattered info into a consistent, high-quality investment memo. | 面向 AI Agent 与人工分析师的公司投研框架,用 10 大维度系统梳理商业模式、护城河、成长与估值,快速产出结构化投研报告。
openclaw skills install company-investment-researchThis skill provides a systematic framework for conducting comprehensive investment research and due diligence on companies. It structures analysis across 10 critical dimensions to support informed investment decisions.
本技能为公司基本面研究提供一套结构化投研框架,覆盖 10 个关键维度,帮助你从零开始梳理一家公司的商业模式、竞争力、增长与估值,并最终形成一份有逻辑的投资结论。
Use this skill when you need to:
适合在以下场景使用:
"Analyze NVIDIA (NVDA) as an investment using the company-investment-research framework. Follow all 10 dimensions and end with a clear BUY/HOLD/SELL view, including key risks."
「请基于
company-investment-research投研框架,系统分析英伟达(NVIDIA, NVDA)的投资价值,按 10 个维度展开,最后给出 BUY/HOLD/SELL 判断,并列出关键风险。」
"Using the company-investment-research skill, compare NVIDIA vs AMD as AI infrastructure investments. Highlight differences in moat, growth drivers, and valuation, then state which one looks more attractive on a 3–5 year horizon and why."
「使用该投研框架对比分析 NVIDIA 与 AMD 作为 AI 基础设施投资标的的优劣,从护城河、成长驱动、估值三方面重点展开,并给出未来 3–5 年哪个更具吸引力及原因。」
"Run a lightweight version of the company-investment-research framework on Snowflake. Focus on competitive positioning, growth drivers, and valuation to decide whether it deserves full deep-dive research."
「对 Snowflake 做一版简化版投研:重点看竞争地位、成长驱动和估值,判断是否值得投入时间做完整深度研究。」
"Create a 2–3 page investment memo for Tesla using the company-investment-research structure. The target audience is an investment committee; keep language concise but include key numbers and scenarios (base/bull/bear)."
「按照本框架,为特斯拉生成一份 2–3 页的投资备忘录,供投委会讨论使用:语言简洁,但需包含核心数据与基础/乐观/悲观三种情景。」
Below are add-on checklists for specific industries. Use them on top of the 10 core dimensions.
下面是针对特定行业的额外检查项,在 10 大通用维度基础上叠加使用即可。
Typical businesses / 典型业务: 内容平台、电商、社交、短视频、本地生活等。
Extra focus areas / 额外关注点:
User metrics / 用户指标
Engagement & retention / 使用粘性与留存
Monetization model / 变现模式
Unit economics / 单位经济模型
Additional questions / 可直接提问的附加问题:
Prompt 示例: 「针对某互联网平台公司,在 10 大维度基础上,额外重点分析用户增长 & 留存、变现模式与单位经济模型,并结合监管风险给出中长期盈利能力判断。」
Typical businesses / 典型业务: GPU/CPU/ASIC 设计、晶圆制造(Foundry)、封装测试、设备与材料等。
Extra focus areas / 额外关注点:
Position in the value chain / 产业链位置
Technology node & roadmap / 工艺节点与技术路线
End markets & demand drivers / 下游应用与需求驱动
Capacity & supply constraints / 产能与供给约束
Additional questions / 可直接提问的附加问题:
Prompt 示例: 「针对某半导体公司,在 10 大维度基础上,重点补充其在产业链中的位置、主要下游应用结构、工艺/产品路线图以及 AI/汽车电子等结构性需求的暴露度,并评估地缘政治对其业务的潜在影响。」
Typical businesses / 典型业务: 连锁咖啡品牌、连锁茶饮、快餐/休闲餐厅品牌等。
Extra focus areas / 额外关注点:
Store economics / 单店模型
Same-store sales & expansion / 同店增长与扩店节奏
Brand & customer perception / 品牌力与消费者心智
Supply chain & cost structure / 供应链与成本结构
Additional questions / 可直接提问的附加问题:
Prompt 示例: 「针对某连锁咖啡品牌,在 10 大通用维度基础上,重点拆解单店经济模型(投资额、回本周期、毛利结构)、同店增长与扩店策略、品牌定位与复购率,并评估在不同经济周期下的抗压能力。」
When analyzing a company for investment, follow this structured approach to ensure comprehensive coverage:
在分析一家公司时,建议按以下 10 个维度逐一梳理,避免遗漏关键点:
Core Questions:
Search Strategy:
Analysis Approach:
Core Questions:
Search Strategy:
Key Metrics:
Core Questions:
Search Strategy:
Analysis Framework:
Core Questions:
Search Strategy:
Risk Assessment:
Core Questions:
Search Strategy:
Evaluation Criteria:
Core Questions:
Search Strategy:
Analysis Components:
Core Questions:
Search Strategy:
Key Considerations:
Core Questions:
Search Strategy:
Governance Assessment:
Core Questions:
Search Strategy:
Valuation Framework:
Synthesize all analysis into clear recommendation:
If Investment is Recommended:
If Investment is NOT Recommended:
Present findings in a clear, structured format:
# Investment Analysis: [Company Name]
**Date:** [Current Date]
**Analyst:** [Your Name or Agent]
## Executive Summary
[2-3 paragraph overview with key takeaway and recommendation]
## 1. Competitive Positioning
[Findings]
## 2. Technology & Innovation
[Findings]
## 3. Market Position
[Findings]
## 4. Customer Base
[Findings]
## 5. Growth Analysis
[Findings]
## 6. Financial Performance
[Findings]
## 7. International Exposure
[Findings]
## 8. Ownership & Governance
[Findings]
## 9. Valuation
[Findings]
## 10. Investment Recommendation
**Recommendation:** BUY / HOLD / SELL
**Target Price:** [If applicable]
**Investment Thesis:** [Key reasons]
**Key Risks:** [Main concerns]
建议在实际使用中,将上述结构作为 Markdown 模板,一边研究一边填空,最终沉淀为可复用的投研文档库。
Use Multiple Sources / 多源交叉验证
Cross-reference information from company filings, analyst reports, news, and financial databases.
Verify Timeliness / 确保数据新鲜度
Always check dates on data – use the most recent available information.
Quantify When Possible / 尽量量化
Provide specific numbers, percentages, and metrics rather than only qualitative descriptions.
Acknowledge Limitations / 明确假设与局限
Note when information is unavailable or when making assumptions.
Maintain Objectivity / 保持客观
Present both bullish and bearish perspectives; avoid confirmation bias.
Source Attribution / 标注关键来源
Cite sources for key data points, especially financial figures.
This skill focuses on how to think and structure research. For data and filings, pair it with external tools/APIs.
本技能侧重于思考框架与结构化输出,财报与数据建议通过其他工具或脚本获取,然后作为本框架的输入。
示例以美股为主(可按同样思路换成 A 股/港股对应数据源):
Download latest 10-K / 10-Q filings
Use any SEC helper tool or script, for example:
# Example: download the latest 10-K for NVIDIA (NVDA) into ./filings
sec-edgar-downloader company "NVIDIA" \
--form-type 10-K --num 1 --download-folder ./filings
Fetch key financials via an API or Python script
For example, a simple Python entry point:
python scripts/fetch_financials.py --ticker NVDA --out data/nvda.json
The script can query any financial data provider (Yahoo Finance, financial APIs, etc.) and standardize outputs (revenue, margins, key ratios) for later use in this framework.
Use web_fetch for qualitative sections
For business descriptions, risk factors, and management discussion, you can:
1. Download the filing (PDF/HTML)
2. Use web_fetch to extract key sections into markdown
3. Feed those into the 10-dimension analysis
推荐实践:将「数据抓取脚本 + 本投研框架」放在同一项目中,通过 Makefile 或 shell 脚本串联,形成一键跑通的投研流水线(先拉数据,再生成报告草稿)。