Skip to main content

Learning Objectives​

After completing this chapter, participants will be able to:

  • Understand what MCP (Model Context Protocol) is.
  • Understand why MCP is becoming a foundational technology for AI-assisted development.
  • Connect AI systems to external engineering tools.
  • Expose organizational capabilities as reusable AI tools.
  • Integrate Atomic Bomb and other engineering systems through MCP.
  • Move from AI-generated code to AI-executed workflows.

Requirements​

We are going to use the BomberMCP tool in this chapter. Please make sure you have it installed and running before proceeding.

Official package​

Requires Rust 1.72 or later to compile.

git clone https://codeberg.org/ReneKrewinkel/bomber.git
cd bomber
cargo build --release
target/release/bomber
# or
cargo install --path .
bomber
# or just run it directly
cargo run --release

Fallback package (JavaScript version)​

If you don't have Rust installed, you can use the JavaScript version of BomberMCP.

git clone git@github.com:ReneKrewinkel/bomberjs.git
# or
git clone https://github.com/ReneKrewinkel/bomberjs.git

npm install
npm start

In both cases, the tool will start a local HTTP server on port 3042. You can verify that it is running by visiting http://127.0.0.1:3042

Configure tool in copilot​

How to configure BomberMCP in your copilot environment will depend on the specific copilot you are using. Please refer to the documentation of your copilot for instructions on how to add external tools and configure them to communicate with BomberMCP.

  • Inside copilot:
/mcp
# Follow the steps from there