Investment Portfolio Manager MCP Server

Investment Portfolio Manager is a powerful MCP server that connects directly to Claude to provide real-time investment analysis, recommendations, and visualization—all through natural conversation.

Comment

Inspiration

When I used to make investments, I always googled news related to the Stock I am investing in. Now, as advanced investors have complex portfolio composed of stocks, bonds, etc, we need to utilize the AI to manage the portfolio and get insights by examining recent news and updates in the stock market.

What it does

The Investment Portfolio Manager MCP Server serves as a bridge between AI assistants like Claude and the complex world of financial investment management. It enables users to create, analyze, and optimize investment portfolios through natural language conversation. The server provides tools for managing portfolios with stocks and bonds, fetching real-time market data, generating comprehensive portfolio analysis reports, offering personalized investment recommendations, and creating visual representations of portfolio allocations. Users can interact with sophisticated financial management capabilities without leaving their AI assistant's interface, making investment management more accessible and conversational.

How we built it

We built the Investment Portfolio Manager using the Model Context Protocol (MCP), a standardized way for AI assistants to access external data and functionality. The server was developed in Python, leveraging the MCP Python SDK's FastMCP framework for rapid development. We structured the server with a modular architecture, organizing code into distinct components:

  • Core Server Module: Handles MCP protocol implementation, message routing, and connection management
  • Tools: Implements financial functions like portfolio updates, stock data retrieval, and investment analysis
  • Resources: Exposes portfolio data to AI assistants through standardized URI patterns
  • Data Layer: Manages portfolio storage and retrieval using JSON files (with plans to integrate with financial applications)
  • API Integrations: Connects to market data sources for real-time stock information and news

We also implemented flexible transport options, supporting both stdio (for Claude Desktop integration) and SSE (for HTTP-based clients), with robust error handling and graceful shutdown capabilities.

Challenges we ran into

Developing the Investment Portfolio Manager presented several challenges:

  1. Financial Data Complexity: Representing diverse financial instruments (stocks, bonds) and their performance metrics in a standardized format required careful data modeling
  2. Real-time Market Data: Integrating with external financial APIs while managing rate limits, authentication, and data consistency presented technical hurdles
  3. MCP Protocol Implementation: Learning and correctly implementing the Model Context Protocol specification, particularly understanding the distinction between tools, resources, and prompts
  4. Error Handling: Developing robust error handling across asynchronous operations, especially when dealing with external API failures or timeouts
  5. Portfolio Representation: Creating meaningful visualizations and analyses that could be effectively communicated through text-based AI assistant interfaces

Accomplishments that we're proud of

We're particularly proud of several achievements with this project:

  1. Natural Language Financial Interface: Successfully created an intuitive conversational interface for sophisticated financial operations
  2. Modular Architecture: Developed a clean, extensible codebase that separates concerns and allows for easy addition of new financial capabilities
  3. Transport Flexibility: Implemented multiple transport options (stdio and SSE) with automatic retry and error recovery mechanisms
  4. Portfolio Analysis: Created comprehensive portfolio analysis tools that provide meaningful insights and recommendations based on portfolio composition
  5. Visualization Capabilities: Successfully integrated data visualization tools that translate complex financial data into clear graphical representations
  6. Robust Error Handling: Implemented comprehensive error management across all server components, ensuring reliability even when external services fail

What we learned

This project provided valuable learning experiences in several areas:

  1. MCP Development: Gained deep understanding of the Model Context Protocol specification and best practices for server implementation
  2. Financial Data Processing: Learned techniques for handling and analyzing financial market data in real-time applications
  3. AI Assistant Integration: Discovered effective patterns for exposing complex functionality to AI assistants in ways that feel natural in conversation
  4. Asynchronous Programming: Strengthened skills in managing asynchronous operations and concurrency in Python
  5. Service Architecture: Developed expertise in designing robust service architectures with clear separation of concerns
  6. Error Recovery Strategies: Learned advanced techniques for graceful error handling and service recovery in distributed systems

What's next for Investment Portfolio Manager MCP Server

Looking forward, we have several exciting development paths planned:

  1. Financial Application Integration: Moving beyond manual JSON files to integrate with actual banking and brokerage applications for real portfolio data
  2. Enhanced Analytics: Implementing more sophisticated portfolio analysis, including risk assessment, tax implications, and retirement planning
  3. Cryptocurrency Support: Expanding the platform to support cryptocurrency assets alongside traditional stocks and bonds
  4. Automated Rebalancing: Adding tools for automated portfolio rebalancing suggestions based on target allocations
  5. Mobile Integration: Developing mobile application interfaces that complement the AI assistant experience
  6. Multi-user Support: Enhancing the platform to better support financial advisors managing multiple client portfolios
  7. Historical Analysis: Adding capabilities for historical performance tracking and "what-if" scenario modeling
  8. Alternative Investments: Expanding beyond stocks and bonds to include real estate, commodities, and other alternative investments

These enhancements will transform the Investment Portfolio Manager from a powerful portfolio analysis tool into a comprehensive financial management platform accessible through natural conversation.

Built With

Share this project:

Updates

posted an update —

Demo:

For you to Demo, click on the GitHub Repo link.

You need to have Claude installed.

Go to settings and make sure the MCP is turned on.

Open Claude's claude_desktop_config.json file and add in the dependency as provided in the README of the Github link.

Make sure to restart the Claude.

You also need to have the Server running in your terminal.

When running the terminal or installing the requirements, your command "Python" might have to change depending on your machine. For me I had to use "Python 3.11" but this will vary.

If you have any questions on trying this on your local machine, feel free to reach out to me on Discord, @IK.

Log in or sign up for Devpost to join the conversation.