#  Summer Workshop Series: Claude for Teaching, Course Development, and Research 

 



 ##  

  expand\_more  

 
  

 

 ![The logo for Claude AI, a large orange asterisk.](/sites/g/files/omnuum6756/files/2026-04/Claude%20Logo.png)

 

Now that Anthropic accounts are available to all FAS Faculty by request, the Bok Center will be offering a summer workshop series on using Claude for teaching, course development, and research. While most faculty have used the ChatGPT or Claude chat windows, they may not have encountered Claude's other interfaces, nor the advanced workflows they enable. This series will help faculty get started and/or move further with the most advanced tools HUIT has made available.

No previous experience with *any* of these AI tools will be assumed in the first workshop, and no previous experience with Claude Code will be assumed in the second workshop. More experienced users can elect to begin with Day 2 or Day 3.

The four days are structured as a progression: Day 1 introduces Claude and Cowork for absolute beginners; Day 2 moves into Claude Code with built-in commands and context management; Day 3 covers advanced Claude Code tooling, MCPs, skills, hooks, and multi-agent workflows applied to teaching, course development, and research contexts; Day 4 brings it all together in a capstone day developing a course, course content and communicating research.

All sessions are full: **May 18-21 (full), June 1-4 (full), &amp; June 8-11 (full) from 1:00-2:30pm with optional work time 2:30-4pm.** Sessions will be held in-person at the Learning Lab Studio, 50 Church St. 3rd Floor.  
  
Additional sessions will be offered later this summer and will have more information on our website soon.



 

 [ Contact the Bok Center with Questions arrow\_circle\_right ](mailto:bokcenter@fas.harvard.edu) 

 

 

 

 



###    Day 1 — Getting Started with Claude and Cowork (May 18, June 1, and June 8)  expand\_more  

 

**Intended Audience:** Faculty with no prior experience or limited experience with Claude, the Claude desktop app or Claude Cowork. Faculty who already use Claude Cowork or Claude Code regularly may wish to skip this session.

Day 1 will be an orientation to Claude's three interfaces: 1) the chat window most faculty already know, 2) Cowork (an interface that allows you to give Claude read and write access to files on your computer), and 3) Claude Code (a terminal-based tool that takes the abilities of Cowork even further).The majority of the session will be devoted to Cowork. We'll install the desktop app, tour the interface, work through the basics of prompting, and connect Cowork to a folder of course materials or research data so you leave with a working setup on your own machine. We’ll close with a brief explanation of how Cowork and Claude Code differ and a demonstration of what Claude Code and more advanced techniques can enable.



 

 

 



###    Day 2 — Claude Code: Setup, Commands, and Context (May 19, June 2, and June 9)  expand\_more  

 

**Intended Audience:** Faculty who work regularly with the Claude desktop app but have little or no experience with Claude Code, command line tools or code editors. Faculty who are regular users of Claude Code or who are experienced programmers may wish to skip this session

Day 2 will be an on-ramp to Claude Code for faculty ready to move beyond Cowork into a more technical environment. We'll install Claude Code and VS Code, get set up on GitHub, and get comfortable with the built-in commands and settings. We will assume that you have not yet set these tools up, and that you may never have worked in a code editor before. We will help you (both during the session and in an additional 90 minutes of office hours) troubleshoot the many quirky steps of getting command line tools running on your own machine (though we will also have Bok computers available for you to use as well with all the tools preloaded). You will learn how [CLAUDE.md](http://claude.md) files and skills work in the Claude Code environment, how to manage context files in VS Code, and, of course, how and why you might want to write some code (even if you’ve never done it before!). We'll close with a preview of skills to set up Day 3.



 

 

 



###    Day 3 — Advanced Claude Code: Skills, MCPs, Hooks, and Multi-Agent Workflows (May 20, June 3, and June 10)  expand\_more  

**Intended Audience:** Faculty who have their Claude Code development environment set up and working, and are ready for more advanced techniques.

The session is for faculty who have already installed Claude Code and are ready to go beyond the basics. We'll work through skills (folders Claude Code discovers automatically when a task pattern matches), MCP integrations (connections to Google Drive, GitHub, databases, and other services), and custom slash commands and hooks. We'll cover design patterns for structuring a context folder so all of these pieces work together, and end with multi-agent workflows. We will cover when complex agent orchestration is worth the setup, and what kinds of problems it actually solves. We will also discuss how to incorporate third-party AI tools using APIs.

 

 



###    Day 4 — Claude Code: Putting it all together to develop course content and communicate research (May 21, June 4, and June 11)  expand\_more  

 

**Intended Audience:** Faculty who are comfortable with their understanding of Claude Code but are looking for ideas about (and best practices for) what to build with it.

The capstone day. Using everything from the prior sessions — Cowork or Claude Code, a structured CLAUDE.md, skills, MCPs, and where appropriate multi-agent orchestration — we'll move from raw materials (course notes, lecture records, readings, research data) to outputs faculty can actually use:a revised syllabus with dates and policies updated for next year, problem banks based on your past problems, interactive visualizations that explain key course content or communicate your research. After Day 4, faculty will have a rough-but-real artifact in progress, anchored in their own material, and a gallery of more advanced outputs (interactive textbooks, graphical abstracts, scholarly publications, lab websites and scicomm) to choose from for follow-up sessions.