Skip to content
Get packs

Guide · Agent skill

How to use the Income Investment agent skill (income-investment)

Updated · written with AI from the project's own documentation and checked against our catalogue data

The income-investment skill is an agent skill from the AI Berkshire project that gives your AI agent a structured workflow for judging whether a company can keep paying its dividend or distribution, and whether that income is worth holding. It is aimed at investors who want a consistent check before buying, students learning payout and cash-flow analysis, and analysts who want labeled research reports. After this guide you will know what the skill does, how to install it, how to phrase a request so the agent uses it, and what to check before trusting the output.

Income Investment Analysisxbtlin/ai-berkshire · MIT

Quick start

  1. 1Get the ZIP. Download the income-investment ZIP from this page.
  2. 2Install the skill. Unzip it into ~/.claude/skills/ for all projects or into a project's .claude/skills/ folder, or upload the ZIP under Settings → Capabilities → Skills in the Claude apps.
  3. 3Set up the repository. Check out the ai-berkshire repository so tools/, AGENTS.md and skills/ are available to the agent.
  4. 4Confirm Python 3 and shell access. Make sure Python 3 is available for the helper scripts and the agent can run the date command.
  5. 5Ask a natural question. Name a company or ticker and state whether you want a dividend durability check, a yield check or a portfolio-fit view.
  6. 6Review the labelled report. Check the data cutoff date, the evidence quality rating and any INSUFFICIENT DATA or Not calculable notes before relying on the conclusions.

What is an agent skill and what does this one make the agent do?

An agent skill is a folder containing a SKILL.md file of instructions that an AI agent loads when a request matches the skill. This one is named income-investment and is described as AI Berkshire skill: Income Investment: Durable and Opportunistic Distribution Analysis.

Once installed, the agent follows a research workflow to answer one question: can this company produce sufficiently durable and attractive distributable income to justify a portfolio role? The workflow covers distribution history, cash available for distribution, debt, sector-specific measures, and valuation. It separates verified facts from estimates, assumptions and opinions, and it reports missing data instead of guessing.

  • Checks payout history, including increases, holds, cuts and suspensions
  • Traces cash flow coverage, free cash flow payout, interest coverage and net debt
  • Uses sector-specific measures such as FFO and AFFO for REITs, CET1 for banks and NII and NAV for BDCs
  • Values the income stream, including yield versus history and combined price-decline and dividend-cut scenarios
  • Designed for learning and research, not personalized investment advice

Install the income-investment skill

There are two routes. You can download the ZIP from the skill page and unzip it into your personal skills folder at ~/.claude/skills/, or into a project's .claude/skills/ folder if you want the skill limited to one project. Alternatively, in the Claude apps you can upload the ZIP under Settings, then Capabilities, then Skills.

The skill's SKILL.md refers to files in the ai-berkshire repository, including tools/, AGENTS.md and skills/. Run the workflow from a checkout of that repository, because the skill's helper scripts live there.

  • Download the income-investment ZIP from this page.
  • Unzip it into ~/.claude/skills/ for all projects, or into your project's .claude/skills/ for one project.
  • Or upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
  • Make sure you also have the ai-berkshire repository checkout containing tools/, AGENTS.md and skills/.
~/.claude/skills/

What you need before running it

The skill depends on several things besides the instructions file. Check these before you start so the report does not stall partway through.

  • Python 3, to run the tools/financial_rigor.py and tools/report_audit.py helper scripts
  • Shell access, so the agent can run the date command to confirm today's date
  • Web search or document access, to gather financial filings and market data
  • The ai-berkshire repository checkout containing tools/, AGENTS.md and skills/ files

How to trigger the skill with a natural request

The skill accepts a company or ticker, which is required. Every other field is optional. You can use the command form or plain language. Optional fields include mode (new or existing), role (core-income, opportunistic-income or unspecified), quantity, cost_basis, portfolio_weight, target_yield, tax_residence, portfolio_file and horizon.

If you leave out optional values, the skill marks them Unknown or Not calculable and tells you the consequence rather than inventing them.

  • Command form: /income-investment "<company or ticker>" mode=new role=core-income
  • Plain language: Use the income-investment skill to check whether Company X can sustain its dividend.
  • Add portfolio details only if you want a portfolio-fit check, for example a current weight or cost basis.
/income-investment "<company or ticker>" [mode=new|existing] [role=core-income|opportunistic-income|unspecified]

Examples of what to ask

Good requests name the security and state the question. Be specific about whether you want a durability check, a net income figure or a portfolio-fit view.

  • Is the dividend of [company] sustainable given its free cash flow coverage and debt?
  • Analyze [ticker] as a core income holding, mode=new, role=core-income.
  • Check whether [ticker]'s high yield is a yield trap.
  • Calculate the gross annual dividend income for 500 shares of [ticker], without net income.

Understand the output and its limits

The report starts with a data cutoff date, taken from the date command. It labels material statements as Verified fact, Estimate, Assumption or Analytical judgment, and rates evidence quality from A to C. If fundamentals are materially insufficient, the workflow stops at an INSUFFICIENT DATA gate.

Net income is only calculated when tax residence, account type, applicable treaty and confirmed withholding are supplied. Otherwise the report shows gross income and explains why net income cannot be calculated.

  • Expect a data cutoff date in the report header
  • Expect labelled statements and an evidence quality rating
  • Expect the report to say when net income is not calculable

Safety notes: scripts and network access

This skill runs scripts and uses the network. The helper scripts tools/financial_rigor.py and tools/report_audit.py are run with Python 3, and the agent uses web search or documents to gather filings and market data. The skill's output depends on those sources, which may be incomplete or wrong.

Read the scripts in the repository before running them, and review the sources the report cites. The skill is not personalized investment advice, so do not make a real buy or sell decision on its output without independent review.

  • Review tools/ scripts before running them
  • Check cited sources yourself, especially for decision-critical figures
  • Do not treat a single report as a buy or sell instruction

What people use it for

  • Testing whether a company's dividend is covered by free cash flow before adding it to an income portfolio
  • Learning how payout ratios differ between REITs, banks, insurers and utilities
  • Spotting a yield trap by comparing a high current yield with history and a possible dividend cut
  • Producing a consistent, labelled research note with an audit step for a team
  • Checking how a dividend-paying holding fits an existing portfolio's sector and geographic concentration

Frequently asked questions

How do I install the income-investment agent skill?

Download the ZIP from this page and unzip it into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ folder for one project. In the Claude apps you can instead upload the ZIP under Settings → Capabilities → Skills.

What do I need to run the income-investment skill?

You need Python 3 for the tools/financial_rigor.py and tools/report_audit.py scripts, shell access for the date command, and web search or document access for filings and market data. You also need the ai-berkshire repository checkout containing tools/, AGENTS.md and skills/.

Is the income-investment skill free?

The source does not state pricing or licence terms beyond the listed licence, which is MIT. Check the repository licence for the terms that apply to you.

Does the skill give personalized investment advice?

No. The skill is designed for learning and research and is explicitly not personalized investment advice. Its output depends on web and document sources that may be incomplete or wrong.

Why won't the skill calculate my after-tax dividend income?

It refuses to estimate net income without your tax residence, account type, applicable treaty and confirmed withholding treatment. Without those it shows gross income and explains why net income is not calculable.

Can I use this skill for a quick yield lookup?

It is not designed for that. The workflow is long and built for full research reports, so a simple price or yield check is better done another way.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Use income-investment if you want a structured, labelled check on whether a dividend is sustainable, or you are learning how payout and cash-flow analysis works, and you have Python 3 and the ai-berkshire repository available. Pick something else if you need a quick yield lookup, a personalised recommendation, or a precise after-tax figure without supplying your tax details.

Download the skill