Turn Claude Code Into Your Personal Wall Street Analyst

Open a terminal and start Claude Code.

Once Claude Code is running, type:

/marketplace

That opens Claude Code's plugin manager. You will see tabs for Discover, Installed, Marketplaces, and Errors.

The important part is that this command is not a normal chat prompt. It opens the interactive marketplace interface where you can add new plugin sources and install skills.

Inside the plugin manager, go to the Marketplaces tab.

Select Add Marketplace.

Paste this GitHub URL:

https://github.com/anthropics/financial-services

After you add it, Claude Code should show a new marketplace named claude-for-financial-services.

In our run, the marketplace showed 20 available plugins, with several already installed.

Open the financial-services marketplace and choose Browse plugins.

The skills you install will depend on the work you want Claude to do. For a personal market research setup, these are the strongest starting points:

• market-researcher for sector research, industry overviews, competitive landscapes, and peer comparisons
• earnings-reviewer for earnings calls, filings, and model update notes
• equity-research for equity research workflows, earnings analysis, initiation-style reports, and research memos
• financial-analysis for financial modeling, comparable company analysis, DCF-style analysis, LBO-style analysis, and spreadsheet outputs

Use the keyboard controls shown at the bottom of the plugin manager. In the screenshots, Space toggles a plugin, Enter opens or selects, and Esc goes back.

Install only what you need at first. It is easier to test one workflow cleanly than to install everything and forget which skill is doing what.

Once the skills are installed, start with a bounded research request. Do not ask for "the market" in general. Give Claude a sector, ticker list, company, or event.

For a daily market brief, you can use a prompt like this:

Use the market research skill to create a pre-market research brief for this watchlist:
NVDA, MSFT, AAPL, TSLA, AMZN, META, GOOGL.

Use public sources only.
Include:
- Broad market snapshot
- Biggest watchlist movers
- Likely reasons for the moves
- Upcoming earnings or events
- Macro events to watch
- Source links for important claims

Do not give buy or sell recommendations. Separate facts from interpretation and flag uncertainty.

For a spreadsheet-style output, you can ask:

Use the financial analysis skill to create an Excel spreadsheet comparing these companies:
Lumentum, Coherent, and IPG Photonics.

Include operating metrics, valuation multiples, margin comparison, and a short summary of what stands out.
Use public information and cite sources where possible.

In the test run, Claude produced a comparable company analysis spreadsheet with operating statistics, revenue, growth, gross margin, EBITDA, net income, EPS, valuation multiples, and summary statistics.

That is the practical payoff. The plugin setup turns Claude Code into a research workspace that can produce structured files, not just paragraphs.

You can also use the newly installed skills from Claude Code inside the Claude desktop app.

Open Claude desktop, go into Claude Code, and ask for the same type of finance task. The skills you installed give Claude a more specific way to handle the work, especially when you need reports, spreadsheets, or repeatable research formats.

This is useful when you want the workflow to feel less like a terminal session and more like a recurring assistant.

Once the research prompt is reliable, turn it into a routine.

For example, you can create an earnings scanner that runs every Monday morning:

The routine in the screenshot uses this instruction:

Check to see what important earnings reports are coming up this week. Give me a report of guidance and use market research skill to give me context on each company.

You can adapt that into a daily or weekly market update:

Every weekday morning, use the market research skill to create a concise market update for my watchlist.
Include major overnight moves, relevant headlines, upcoming earnings, macro events, and questions I should verify before taking action.
Do not give financial advice or trade recommendations.

Start with a weekly routine before you automate anything daily. Weekly updates are easier to review for quality, and they give you time to catch weak sources, vague summaries, or missing citations.