Give Your AI Agent a Debugger: C-SPY, MCP, and Python in the Modern IAR Workflow

  • Europe & Asia - Nov 19th
  • 10 AM CET
  • 45 minutes

You’ve tried AI coding assistants. They write firmware fast, then hit a wall: they can’t see your registers, your memory, or your call stack, and they guess at toolchain options. In embedded work, a wrong guess costs an afternoon.

Now imagine your agent reading the IAR documentation, starting a C-SPY session on real hardware, stopping at a breakpoint, and finding a bug you can’t see in the source. Then imagine it turns that finding into a test that runs in CI every night. With MCP (Model Context Protocol) servers and a Python package, it can.

That’s what we’ll show, using the modern IAR workflow, then take your questions live. No board required to try it yourself.

This session shows what the IAR Platform now enables when you give your AI agent a debugger. You will see: 

  • An AI agent grounded in the official IAR documentation through an MCP server, with links back to the source

  • An agent debugging a running target through the C-SPY MCP server, alongside IAR debugger features like runtime error checking and trace

  • A Python debug script turned into a repeatable pytest-based firmware test for your laptop and CI

  • How to try the same setup on your own project, with an IAR cloud license

The agent drives. You stay the engineer, and you decide what ships.

Register now to learn how to take advantage of AI agents on your next embedded project.

New to the IAR Platform? Check this demo.

Register for webinar

Related content

Embedded in Industries

medical diagnostics imaging

IAR support across all  medical applications

yellow_industry_robot_adobestock_213480015

IAR support across all  industrial automation applications