Turn course evaluations into revisions with Claude Code

A professor fed evaluation PDFs to Claude Code, used the results to guide edits, and repeated the process for three classes.

Done with Claude Codeยท Intermediate

What they did
Nick Feamster uploaded a course-evaluation PDF to Claude Code and asked it to extract the recurring themes. He copied those themes into a GitHub issue as a checklist, then let agents carry out the updates: they added rubrics to assignments, made exam policies uniform, and moved around two dozen slide decks into Quarto, handling one deck per agent in batches of about six. Before letting the process scale out, he checked a reference deck that he built by hand.
How it went
His security class was updated with rubrics, a syllabus section about communication, aligned exam policies, and refreshed slides. He then reused the same playbook for the ML systems and censorship classes. An agent also located an old final exam and rebuilt it from blank source files.
Tips
He reviewed diffs throughout and required confirmation before any commit or push. He also treated web-based claims as items to verify, not as facts to accept automatically.

Try it yourself with Claude Code

I have student evaluation PDFs in [folder] for [course name]. Extract the repeated themes and organize them into praise, problems, and specific changes I should make. Then draft exact edits for my teaching materials in [repo or folder], and apply them on a new branch only after you show me the plan and I approve it. Stop and ask before any commit, push, delete, or submit action, and leave out student-identifying details.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.