AI for Research — demos, slides, and guides

Getting started start here

New to this? Setup is a one-time thing. Install Claude Code, add the Open Science Skills, and point it at a project. The walkthrough takes it step by step, with nothing assumed.

Set up Claude Code and the skills →

Install the command-line AI tool that runs the skills, sign in, add the Open Science Skills, and confirm they loaded. About five minutes, no prior setup assumed.

step-by-step

Demos run them yourself

Small projects you clone and run end to end, each with sample files, the exact prompts, and expected output.

Reference & source-claim checking →

Catch fabricated or malformed citations against your reference list alone. Then, with a small set of your sources on hand, check whether each one actually supports the claim it backs. Runs offline on a short sample article with planted errors.

citation-check · fact-check · research-repo

More demos to come.

Lectures & slides teaching

Slide decks and notes from talks and workshops. New decks are added to the lectures folder as they are given.

From Pixels to Patterns →

A talk on the methods behind the Pixels and Patterns project: computer vision and language models in empirical social science and the digital humanities.

slides

More lectures & slides to come.