How to use the Dividend Stock Analysis agent skill

What an agent skill is and what this one makes the agent do

An agent skill is a folder containing a SKILL.md file with instructions that an AI agent loads when a request matches the skill's description. This skill is named dividend-analysis and is in the analysis category. It covers income, dividend-growth and shareholder-return strategies.

Once it is active, the agent works through a fixed set of questions for each company: what the current cash yield is and whether it is normal for the sector, whether earnings and cash flow cover the payout, whether the balance sheet can absorb a down cycle, how management has treated the dividend across cycles, and whether valuation still leaves room for total return. It also explains ex-dividend date mechanics and frames expected total return as scenarios rather than guarantees.

• Coverage checks: earnings payout, free-cash-flow payout and operating cash flow coverage
• Dividend growth compared with EPS and FCF growth
• Balance sheet checks: cash, net debt to EBITDA, interest coverage, debt maturities
• Yield-trap checklist covering yield relative to history and peers, payout ratios, declining earnings and rising leverage
• Scenario-based total return framing

Install the dividend-analysis skill

Download the ZIP file for the skill from the download link on this page, then unzip it. Place the resulting dividend-analysis folder in one of two locations: your personal skills folder at ~/.claude/skills/, or a project's .claude/skills/ folder if you want the skill only for that project.

If you use the Claude apps instead of a folder on your computer, you can upload the skill under Settings → Capabilities → Skills.

• Personal use across projects: ~/.claude/skills/dividend-analysis/
• Single project only: the .claude/skills/ folder inside that project
• Claude apps: Settings → Capabilities → Skills, then upload the skill

Set up the data the skill needs

The skill does not contain financial data. It needs figures for the company you ask about, which you can paste in yourself or which the agent can fetch through the data sources named in the skill file. The file lists dividends, net income, operating cash flow, free cash flow, debt and EBITDA as the fields it relies on.

For A-shares the skill names Tushare, which probably needs an API key. The file names the Tushare fields but does not explain how to set it up, so check Tushare's own documentation for that step. For US and Hong Kong stocks the skill names yfinance, a Python library. The skill file does not give install commands for either library.

• Supply the numbers yourself if you do not want the agent to fetch them
• If live data is unavailable, the skill tells the agent to state the limitation and give a template instead of inventing figures

Ask for dividend analysis in natural language

You do not need a special command. The skill is meant to be triggered when you ask about dividend stocks, income portfolios, dividend growth, high-yield screening, payout safety, ex-dividend dates, or whether a dividend is sustainable. Name the company or ticker and say what you want to know.

Good requests state the goal. For example, ask whether a company's dividend is covered by free cash flow, or whether a 9% yield is a yield trap. Mention the strategy if it matters, such as income, dividend growth or high-yield quality, because the skill treats those differently.

• Is the dividend at [company] covered by free cash flow over the last few years?
• Check [ticker] for yield-trap risk and tell me what would break the dividend.
• Compare the payout quality of these three dividend stocks using the same checklist.
• How does the ex-dividend date work for [ticker], and is buying before it worth it?

Read the output and its limits

For a company, the skill is designed to return a verdict (sustainable, watchlist or yield-trap risk), a metrics table with interpretations, what supports the dividend, what could break it, and base, downside and upside scenarios. It ends with a note that the output is investment research, not live trading advice.

The results are only as good as the data behind them. Stale or incomplete figures produce weaker conclusions, so check the inputs before relying on a verdict.

Safety notes: scripts and network access

This skill does not run scripts. Its code blocks are formulas for the agent to apply, not programs you need to execute. The skill does use the network, because it can fetch data through Tushare and yfinance when those are set up.

The skill is an analytical framework, not personalised investment advice. It does not give tax planning advice and mentions withholding tax and taxes only in general terms.

• Scripts: none
• Network: yes, when fetching data from Tushare or yfinance
• Not for guaranteed or personalised advice
• Not for non-dividend stocks, since the checks assume dividends are paid
• Not for jurisdiction-specific tax rules

Tips for better results

Use the same request wording across companies so the comparisons are consistent. Ask about one company at a time for the deepest checks, and separate ordinary dividends from special dividends in your question, since the skill treats them differently.

For REITs, utilities, banks, MLPs and insurers, say the sector in your request. The skill adapts the payout metric to the sector, for example AFFO payout for REITs.

• Separate regular and special dividends
• Say the sector for REITs, banks, utilities, MLPs and insurers
• For ADRs and cross-listed shares, mention withholding tax and currency, since the skill accounts for them