Agentic Coding for Social Impact

A hands-on workshop on building working tools with AI coding agents such as Claude Code. No programming experience is needed. After you register, there are two steps before the workshop.

Step 1. Pre-workshop survey

About 15 minutes. The survey is part of the workshop evaluation and asks about your background, your views on technology and AI, and a short quiz. Instead of your name, you create a personal code from your two initials and the year you were born (for example, JS1998) and enter the same code in the survey at the end of the workshop.

Step 2. Pre-work

About 50 minutes in total, in any order. With the setup done in advance, the session can start with design and building.

Part	Time	Result
Setup guide	20 min	The Claude desktop app, a one-page test site that shows the setup works, a GitHub account, and optionally a Vercel account
Coding agent basics	15 min	What a coding agent does, the desktop app, permission requests, result checks and the data rule, with optional official videos
Problem worksheet	15 min	A real problem from your placement or work in four short answers; the workshop starts from it

Checklist

Tick each item as you finish it. The ticks are saved in this browser only.

• Pre-workshop survey submitted
• Claude account with Claude Code access
• Claude desktop app installed; the Code tab opens
• Test site opens in the browser and shows your name
• GitHub account created and GitHub Education application submitted, if eligible
• Coding agent basics read
• Problem worksheet downloaded as a Word file

Workshop day

• Your laptop, fully charged, and its charger.
• Your problem worksheet as a Word file on your laptop (printing is optional).
• Only records you make up; no client, placement or employer data.

Instructor: Nari Yoo. Facilitator: Jes Kane. Sponsored by the Program Evaluation Group. For questions, see the FAQ or reply to the workshop email.