AI-assisted research programming
UMD graduate development economics · Fall 2026
This site holds the materials for the AI and programming side of the semester-long LSMS project — a guest module by Emily Beam (University of Vermont). Everything here is built to outlast the course: the setup, the habits, and the code you write are yours to keep.
Start here
| If you’re… | Go to |
|---|---|
| Setting up before the semester | Setup guide — do this before the first class (August 31). Start early: the GitHub student-benefits verification can take a few days. |
| Looking for the lecture slides | Posted here before August 31 |
| Looking for starting-point files | Templates — config pattern, cleaning/analysis skeletons, AGENTS.md house rules |
| Looking for cheat sheets and links | Materials — Git cheat sheet, research decisions log, links worth your time |
The workflow
You’ll do the semester project in VS Code, with an AI agent working beside your code and your files in your own GitHub repository. That’s more setup than opening a chat window — but it’s the version where you can actually see what the AI did to your analysis, and where the code survives the semester.
The loop to remember:
agent writes → you review → it runs → you check the output → you commit
Reuse
Materials are CC-BY 4.0 and draw on Thinking with Agents, an open collection of AI modules for economists. Instructors: take whatever’s useful — everything is plain markdown and LaTeX in the repo.