Skip to content

Quick Start: IDL

IDL Agent empowers you to autonomously write, run, and troubleshoot IDL code, simplifying developer workflows and accelerating discovery.

IDL Agent works by providing a set of Model Context Protocol (MCP) tools and instructions for GitHub Copilot.

Setup

Information

IDL Agent setup is the same as ENVI Agent.

Requirements:

Helpful Tips

Automatic Startup: IDL starts automatically when needed, no manual intervention required.

Creating IDL Notebooks: Explicitly request an "IDL Notebook" to ensure you get the native VSCode implementation rather than a Python/Jupyter notebook. Native IDL Notebooks require no additional setup or configuration.

Example Questions

Here's some requests for what kinds of questions IDL Agent can help with:

Can you execute this IDL code for me?

print, 'Hello world!'
Can you tell me more about the HTTPRequest class in IDL?
Can you create an IDL Notebook for me?

I would like it to create and plot sinusoidal data.
Can you write an IDL routine for me that performs non-maxima-suppression?

I would like it to be a function and have a vectorized/optimized approach
that goes through the bounding boxes one time (so it is more linear for performance)

Attaching Context

GitHub Copilot can incorporate various sources to better understand your code and provide more relevant answers. Click the paperclip icon in your chat to explore available options.

Ways to Add Context:

  • Highlight code in an editor, it attaches automatically
  • Attach multiple PRO files to explain or compare code
  • Reference specific files or folders for targeted assistance

Licensed under MIT.