Day 1Monday, July 20, 2026

Course kickoff and how LLMs and agents work

9:00–9:30 AM

Course Introduction

30 min
  • Welcome and introductions
  • Course objectives
  • Course timeline
  • Projects and project structure
  • Logistics, accounts, and access
9:30–9:45 AM

Setting Up the AI Course Assistant (Aca)

15 min
  • Meet Aca — a course-aware assistant that coaches you, not one that does the work for you
  • Confirm your Claude account and that the Aca skill is provisioned
  • Run `Aca init` to build your personal profile
  • Where to learn more and get help during the course
9:50–10:40 AM

How LLMs Work

50 min
  • A working mental model of large language models
  • Tokens, context windows, and sampling
  • Capabilities and failure modes
10:45–11:35 AM

From Models to Agents

50 min
  • What makes a system “agentic”
  • Tools, loops, and environments
  • Respecting Agent capabilities, boundaries, and constraints
11:40 AM–12:00 PM

Introduction to Context Engineering

20 min
  • What do you know that the agent needs to know
  • Structuring context for better results
  • Best practices for effective prompting
1:00–2:00 PM

Team Project Time & Office Hours

60 min
  • Team project work time
  • Office hours — get your questions answered
  • Mentoring and hands-on practice

Day 2Tuesday, July 21, 2026

Working with Claude

9:00–9:30 AM

Introduction to Context Engineering (continued)

30 min
  • What do you know that the agent needs to know
  • Structuring context for better results
  • Best practices for effective prompting
9:40–10:30 AM

Meet the Claude Family

50 min
  • Introduction to the Claude Model Family
  • Introduction to Harnesses: Claude, Claude Cowork, Claude Design, and Claude Code
  • When to use each
  • Managing local and remote workflows
10:40–11:30 AM

Context Engineering with Claude & Claude Cowork

50 min
  • Getting information into the system
  • Structuring context and prompts for better results
  • Modes and Models, Oh My!
  • Retaining and using artifacts effectively
11:40 AM–12:00 PM

Hands-On: Produce a Project Report

20 min
  • Guided exercise: a polished report with Claude Cowork
  • Iterating and refining results
  • Validating and editing the output
1:00–2:00 PM

Team Project Time & Office Hours

60 min
  • Team project work time
  • Office hours — get your questions answered
  • Mentoring and hands-on practice

Day 3Wednesday, July 22, 2026

Supercharging your Agents: Skills, Tools, Subagents, and Workflows

9:00–9:30 AM

Context Engineering with Claude & Claude Cowork (continued)

30 min
  • Getting information into the system
  • Structuring context and prompts for better results
  • Modes and Models, Oh My!
  • Retaining and using artifacts effectively
9:40–10:30 AM

Using Skills and Tools

50 min
  • Introduction to Skills, tools, workflows, subagents, and MCP
  • When to use each
  • Assessing quality, safety, and reliability
10:40–11:30 AM

Building Skills

50 min
  • Building Skills with Skills
  • Packaging, shipping, and sharing
  • Build and share your own skill
11:40 AM–12:00 PM

Deep Research

20 min
  • Deep research agents
  • Strategies and techniques for effective deep research
  • Leveraging research results
1:00–2:00 PM

Team Project Time & Office Hours

60 min
  • Team project work time
  • Office hours — get your questions answered
  • Mentoring and hands-on practice

Day 4Thursday, July 23, 2026

Scaling up: Claude Code

9:00–9:30 AM

Deep Research (recap)

30 min
  • Deep research agents
  • Strategies and techniques for effective deep research
  • Leveraging research results
9:40–10:10 AM

Skill Exercise: Building & Sharing Simple Skills

30 min
  • Build a simple skill
  • Package and share it
  • Try out a classmate's skill
10:20–10:50 AM

Getting to Know Claude Code

30 min
  • Agents in the terminal
  • Workspace structure and CLAUDE.md
  • Slash commands, skills, and hooks
  • Choosing models and managing resource use
11:00–11:30 AM

The Biology of LLMs

30 min
  • Evolution of large language models
  • Coevolution of models, tools, and the humans who use them
  • Emergent behavior in these systems
11:30 AM–12:00 PM

Ethics, Quality & Failure Modes

30 min
  • Ethical considerations in AI-assisted research
  • Ensuring scientific quality and reproducibility
  • Failure modes in software engineering, data analysis, and writing/research
  • Preempting, diagnosing, and managing failures
  • Discernment and critical evaluation of AI-generated content
1:00–2:00 PM

Team Project Time & Office Hours

60 min
  • Team project work time
  • Office hours — get your questions answered
  • Mentoring and hands-on practice

Day 5Friday, July 24, 2026

Claude Code and wrapping up Week 1

9:00–10:20 AM

Getting Comfortable with Claude Code

80 min
  • Agents in the terminal
  • Workspace structure
  • Context and Claude.md
  • Multi-agent coordination and communication
  • Leveraging Claude Code Slash Commands, Skills, and Hooks
  • Monitoring resource use
  • Choosing models and approaches for effective and efficient resource use
  • Managing and monitoring parallel sessions and agents
10:30–11:20 AM

Team Working with Claude

50 min
  • Applying Claude across your team's workflow
  • Collaboration patterns and handoffs
  • Hands-on team session with instructors on hand
ContentPDFZoomResourcessoonRecordingsoon
11:30 AM–12:00 PM

Week 1 Wrap-Up: State of the Art & Where to Go Next

30 min
  • State of the art in LLMs and agents
  • Where to go from here
  • Staying abreast of new developments
ContentPDFZoomResourcessoonRecordingsoon
Daily rhythm. Morning sessions run 50 minutes with 10-minute breaks between them; all times are Eastern. Each day closes with a 1:00–2:00 PM team project time and office hours session — time for teams to work together, for questions, and for mentoring and practice. It is open to everyone whether or not you are on a team; team participation is optional.