Project setup

Project Overview

• Building a CLI-based chatbot to understand how MCP clients and servers work together
• Users can interact with a collection of in-memory documents (fake data)
• Project includes both a custom MCP client and custom MCP server in one codebase

MCP Server Implementation

• Two tools will be implemented:
• Read document contents
• Update document contents
• Documents are stored in memory only (not persisted to disk)

Important Context

• In real projects, developers typically build either a client OR a server, not both
• MCP servers are distributed to allow other developers to access services
• MCP clients connect to external servers built by other engineers
• This project builds both for educational purposes to demonstrate how they interact

Project Setup Steps

• Download the CLI_project. zip starter code file
• Open the project directory in a code editor
• Read the readme. md file for setup instructions
• Add API key to the . env file
• Install dependencies using UV or pip
• Run the project with uv run main. py (or python main. py without UV)
• A chat prompt should appear; test with a simple query like "what's one plus one"

Next Steps

• After setup is complete, begin adding new features to the application

Takeaways

• Building a CLI chatbot project that includes both a custom MCP client and server to learn how they work together
• MCP server will implement two tools: read document contents and update document contents (stored in memory only)
• Typically developers build either a client OR server, not both — this project does both for educational purposes
• Setup requires downloading starter code, adding API key to . env, installing dependencies with UV or pip, and running with uv run main. py
• After setup, you'll test the chatbot with a simple query before adding new features