Deep company analysis

Skill desmondc9/agent-skills/deep-company-analysis

Claude Code agent skills: multi-expert stock analysis, markdown-to-docx

Install

npx -y skills add desmondc9/agent-skills --skill deep-company-analysis

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generate a structured deep-dive research report on any company, stock, or crypto project from both growth and value investing perspectives. Use when the user asks to '深度分析', '调研', 'research', 'analyze' a specific company, stock, or token — or mentions evaluating an investment target with a comprehensive framework. Triggers include: '分析XX公司', '调研XX股票', '深度研究', '公司基本面分析', '企业调研报告', 'crypto project analysis'. Do NOT use for simple price checks, real-time trading signals, portfolio optimization, or tax advice.

SKILL.md

3.1 KB, as published. Nobody here has run it

深度企业分析 (Deep Company Analysis)

Quick start

Ask the user for the target company name / stock ticker / token symbol (if not already provided).
Read REFERENCE.md — this is the full 8-dimension analysis framework.
Perform web research to gather quantitative data (financials, market share, on-chain metrics, etc.).
Generate the report in Simplified Chinese strictly following the output order and format in REFERENCE.md.
Output the complete report directly to the user (no intermediate files required unless requested).

Workflow

Step 1 — Identify target

Resolve company name, ticker, or token symbol.
Determine asset type: public stock / private company / crypto/Web3 project.
Note the founding date (to determine available historical data range).

Step 2 — Research & fill framework

Use WebSearch/WebFetch to gather data for each dimension in REFERENCE.md.
For every quantitative claim, annotate the source (财报/招股书/第三方机构/链上数据/行业白皮书).
Mark speculative content with 【推测】.
If a dimension is not applicable (e.g., a crypto project has no "debt-to-asset ratio"), explicitly label it: 【不适用】原因：...

Step 3 — Generate report

Output order is mandatory: Executive Summary → Dimensions 1–7 → Crypto Appendix (if applicable).
Use ### headers for each sub-dimension.
Present time-series quantitative data in tables.
After each main dimension, append a 200-word "维度核心摘要".
End with a "综合观察清单" of 5–8 trackable indicators/events for the next 12 months.

Step 4 — Compliance check

Verify: no "买入/卖出/持有" recommendations appear.
Verify: all currency values are explicitly labeled (USD / CNY / HKD / etc.).
Verify: competitive benchmarks (vs peers and industry median) are included for key metrics.
Verify: 3-year, 5-year, and 10-year time spans are covered (or noted if insufficient history).

Key rules

Language: Simplified Chinese throughout. English proper nouns, tickers, and metric abbreviations (PE, PB, ROE, FCF, etc.) may remain in original form.
No investment advice: Only facts, logic, and evidence. No price targets or action recommendations.
Progressive disclosure: If the user only wants a specific dimension, read REFERENCE.md and extract only that section while keeping cross-references intact.

Reference

Full framework: REFERENCE.md