Data Science Special Topics · NC State University Libraries

Agentic Coding with Claude

Thursday, October 15, 2026, 2:00–3:30 pm · South Learning Lab 3430, D. H. Hill Jr. Library

AI agents can read your files, run code and fix their own mistakes. The skill now is setting one up, steering it, and checking its work. In 90 minutes you’ll install Claude Code, hand it a messy spreadsheet, and watch it plan, build and check a dashboard, including the moment it goes wrong and how you catch it.

What you’ll leave with

• Claude Code installed and signed in on your laptop
• A working dashboard, built by Claude from a messy spreadsheet
• Your plan checked against the boss’s note, before any code was written
• The dashboard’s key number checked against the official answer
• A CLAUDE.md that remembers your rule, and a /house-style skill you can reuse
• A link to your dashboard you can share (Pro plan or higher), or a screenshot
• Something of your own, built in the last ten minutes

Before you come

1. Check your plan. Following along needs a Claude Pro, Max, Team or Enterprise plan, or a Console account. The free plan doesn’t include Claude Code. No paid plan? Come anyway: you can watch, or pair up and be the one who checks the agent’s work.
2. Install Claude Code with the setup guide. It takes about 10 minutes, and we’ll do it together at the start too.
3. Bring a laptop (Mac or Windows) and its charger.

The materials

SlidesThe live session, step by step Today’s activityStudy-room bookings: every prompt, ready to copy On your ownCampus café sales: a second activity to try after the session Cheat sheetCommands, keys, modes and habits on one page Finished examplesWhat the activities produce, to compare with yours Setup guideInstall, sign in, and fix common problems

Watch the whole activity in two and a half minutes

A captioned walkthrough of the study-room activity, built from a rehearsal run. The terminal screens are simplified re-creations; the dashboards are the real output. There’s no sound; the captions tell the story, and every step is also written out on the activity page.

Today’s plan

Time	What we’ll do
2:00	What makes an agent different from a chatbot
2:08	Install Claude Code together, open the starter folder, and set up a status line
2:25	The loop on a real task: read, plan, act, with memory and skills explained while it builds
2:48	Check it: one number you trust, and make Claude prove it
2:57	Make it stick: project memory and a skill of your own
3:08	Undo a bad change, and share your dashboard
3:14	Make it yours: build something you want
3:23	Habits for using agents responsibly

Get help

• Questions about the workshop: lib-datascienceservices@ncsu.edu
• Help with access, accounts or setup, before or after: go.ncsu.edu/getdatahelp
• The official Claude Code docs

All data in these activities is synthetic, made up for teaching. Content is licensed CC BY 4.0, and code MIT. Materials checked against Claude Code 2.1.299 on October 14, 2026.