By the end of the lab, you will be able to:
- Configure Claude Code for structured, project-based development
- Define persistent project context using CLAUDE.md
- Create modular, reusable agent behaviors with Skills.md
- Translate workflow requirements into agentic logic
- Connect agents to external APIs and services using MCP
- Apply structured patterns for error handling and extensibility
- Evaluate basic production considerations: security, monitoring, and cost
- Adapt the patterns learned to evolving AI tooling environments
This lab is intended for developers and technical leads who already write code and want to move beyond ad-hoc AI usage into structured agent design. Participants typically:
- Have explored Claude or other LLM tools and want a more disciplined development approach
- Are responsible for integrating AI workflows into existing repositories
- Need to evaluate Claude Code within a team or production context
- Want hands-on experience with MCP integrations before deploying them in live systems
Basic programming experience is required (Python preferred). No prior experience with Claude Code is necessary.
This lab is also available for private team delivery — contact us for customized engagements.
Do I need prior experience with Claude Code?
No. The lab assumes programming experience but introduces Claude Code from first principles, with structured setup and guided configuration.
What will I build?
You will design and implement a working research workflow agent that accepts queries, retrieves and synthesizes information, integrates external tools, and handles errors within a structured Claude Code project.
Is this mostly lecture or hands-on?
Primarily hands-on. Concept briefings are short and immediately followed by applied exercises. You will work in a real code environment throughout the session.
What makes this different from on-demand courses?
This lab is live, instructor-led, and small-group. You receive direct feedback while working in realistic repositories and applied workflows, rather than following pre-recorded demonstrations.
Will I receive a credential upon completion?
Yes. Participants who complete the lab receive a Caltech CTME Letter of Completion and 0.4 continuing education units (CEUs), documenting your participation for professional development records and employer reimbursement purposes.
What is CLAUDE.md?
CLAUDE.md defines persistent project context—goals, constraints, conventions—so each Claude session operates within a structured environment.
What is Skills.md?
Skills.md defines modular, reusable capabilities that make agent behaviors maintainable rather than one-off scripts. This makes it easy to build, test, and evolve agent capabilities independently and without rewriting core logic.
What is MCP?
Model Context Protocol (MCP) is an open standard adopted across the industry, including by OpenAI, Google, and Microsoft. It is for connecting AI agents to external tools, APIs, and services. The lab includes a guided integration exercise.
What do I need to bring?
A laptop with internet access and a code editor. Pre-session setup instructions will be provided in advance.
Will I receive materials afterward?
Yes. Participants receive code templates, reference patterns, and lab materials for continued development.
Can this be delivered for a private team?
Yes. The lab can be customized for engineering teams integrating Claude Code into their existing repositories and workflows.
