AI-Augmented Development & Research Tutorial

AI-Augmented Development & Research

Who this is for: Scientists, research software engineers, and curious newcomers who want practical, role-specific guidance for working with AI coding assistants.

About this tutorial

AI assistants like Claude can accelerate research and software work — but only when used with the right habits. These cheat sheets distill role-specific workflows, prompting patterns, and guardrails into a few scannable pages you can keep next to your keyboard.

Choose a cheat sheet

A prompt to start with

New to this? Copy a prompt like the one below into Claude, then adapt it to your task. Be specific about your role, your goal, and the constraints.

You are helping a research scientist. I have a CSV of daily station
temperatures (columns: date, station_id, tmax_c). I want to compute
monthly means per station and flag months with >5 missing days.

- Walk me through the approach before writing code.
- Use pandas; explain each step briefly.
- Point out any assumptions you are making.
Tip Ask the model to explain its plan before it writes code. It is faster to correct a plan than to debug a wrong implementation.
Caution AI output can be confidently wrong. Verify numbers, citations, and code against trusted sources before relying on them — especially for published results.