Skip to main content
3 min read

Building an MCP-powered Financial Analyst

100% local.

👉

In this chapter, we'll build a financial analyst that connects to your Cursor/Claude and answers finance-related queries.

The video below depicts a quick demo of what we're building!

0:00
/0:32

Tech stack:

  • CrewAI for multi-agent orchestration.
  • Ollama to locally serve DeepSeek-R1 LLM.
  • Cursor as the MCP host.

System overview:

  • The user submits a query.
  • The MCP agent kicks off the financial analyst crew.
  • The Crew conducts research and creates an executable script.
  • The agent runs the script to generate an analysis plot.

​You can find the code in this GitHub repo →​

Let's build it!

Setup LLM

We will use Deepseek-R1 as the LLM, served locally using Ollama.

Let’s set up the Crew now.

Query parser Agent

This agent accepts a natural language query and extracts structured output using Pydantic.

This guarantees clean and structured inputs for further processing!

Code Writer Agent

This agent writes Python code to visualize stock data using Pandas, Matplotlib, and Yahoo Finance libraries.

Code Executor Agent

This agent reviews and executes the generated Python code for stock data visualization.

It uses the Code Interpreter tool by CrewAI to execute the code in a secure sandbox environment.

Setup Crew and Kickoff

Once we have our agents and their tasks defined, we set up and kick off our financial analysis crew to get the result shown below!

Create MCP Server

Now, we encapsulate our financial analyst within an MCP tool and add two more tools to enhance the user experience.

  • save_code -> Saves generated code to local directory
  • run_code_and_show_plot -> Executes the code and generates a plot

Integrate MCP server with Cursor

Go to: File → Preferences → Cursor Settings → MCP → Add new global MCP server.

In the JSON file, add what’s shown below 👇

Done! Our financial analyst MCP server is live and connected to Cursor!

You can chat with it about stock data, ask it to create plots, etc. The video at the top gives you a walk-through.

​You can find the code in this GitHub repo →​

Thanks for reading!

Published on Sep 5, 2025