📝 Creación de MCP Servers y su uso con GitHub Copilot en modo A…

What is 📝 Creación de MCP Servers y su uso con GitHub Copilot en modo Agente?

This repository contains materials, examples, and an implementation of an MCP (Model Context Protocol) server for managing sticky notes. It is used in a talk about integrating custom MCP servers with GitHub Copilot in agent mode and allows AI models to create, read, and summarize notes stored in a text file.

How to use 📝 Creación de MCP Servers y su uso con GitHub Copilot en modo Agente?

Install Python 3.10+, Node.js 18+, uv, and GitHub Copilot (agent mode) in VS Code. Clone the repository, create a uv virtual environment, and install dependencies (mcp[cli], pydantic, python-dotenv, pyyaml). Configure the MCP server in VS Code’s settings.json with a command like uv run --with mcp[cli] mcp run C:\path\main.py. For Claude AI, run mcp install path/to/main.py and edit claude_desktop_config.json. Interact with the server using commands such as add_note, read_notes, and get_latest_note.

Key features of 📝 Creación de MCP Servers y su uso con GitHub Copilot en modo Agente

• Implements an MCP server with tools, resources, and prompts.
• Tools: add_note (add a note) and read_notes (list all notes).
• Resource: notes://latest (retrieve the most recent note).
• Prompt: note_summary_prompt (generates instructions for summarizing notes).
• Data is stored in a plain text file (sticky_notes.txt).

Use cases of 📝 Creación de MCP Servers y su uso con GitHub Copilot en modo Agente

• Manage sticky notes through an AI assistant in VS Code or Claude AI.
• Demonstrate how to create and integrate a custom MCP server with GitHub Copilot in agent mode.
• Show the MCP architecture (tools, resources, prompts) using a simple note-taking example.
• Provide a starting point for developers to build their own MCP servers.

FAQ from 📝 Creación de MCP Servers y su uso con GitHub Copilot en modo Agente

What is the purpose of this repository?

It contains a talk’s materials and a working MCP server for sticky notes, used to teach developers how to create and integrate MCP servers with GitHub Copilot and Claude AI.

What are the runtime requirements?

Python 3.10+, Node.js 18+, uv (package manager), a GitHub account with Copilot agent access, and VS Code with the MCP extension. Claude AI is optional for demonstrations.

How do I fix a "Port in use" error on port 6277?

Identify the process using netstat -ano | findstr :6277 and terminate it with taskkill /F /PID <PID>.

Where are the notes stored?

All notes are saved in a file named sticky_notes.txt inside the project directory.

How do I activate the virtual environment if it fails?

If using conda and uv together, first run conda deactivate, then activate the uv environment with .venv\Scripts\Activate.ps1 (PowerShell) or source .venv/Scripts/activate (Bash).