GitHub Agentic Workflows Workshop

Build your first AI-powered GitHub workflow in about 2 hours.

• Who is this for: Developers comfortable with GitHub who want to build practical AI-powered automations. No prior AI tooling experience required.
• What you'll learn: How agentic workflows differ from classic GitHub Actions, how to use the gh-aw CLI, and how to write workflow prompts in Markdown.
• What you'll build: A scheduled, AI-driven daily repository status report that runs automatically in GitHub Actions — no clicks required.
• Prerequisites: A free GitHub account and basic familiarity with repositories, commits, and the terminal. No GitHub Actions experience is assumed.
• How long: About 2 hours end-to-end.

How to start

Open the workshop folder and follow the steps in order, starting at Step 0:

➡️ workshop/00-welcome.md

Each step is short (5–10 minutes). Checkboxes at the end of each step tell you exactly what to verify before moving on.

Workshop curriculum

#	Title
0	Welcome — What We'll Build
1	What You Need Before We Start
2a	Adventure A: Set Up a Codespace
2b	Adventure B: Set Up Your Local Terminal
3	Create Your Practice Repository
4	What Are GitHub Actions?
5	What Are Agentic Workflows?
6	Install the gh-aw CLI Extension
7	Write Your First Agentic Workflow
8	Run and Watch Your Workflow
9	Reading Workflow Output
10	Design: Your Daily Repo Status Report
11	Build: Daily Repo Status Workflow
12	Test and Improve Your Workflow
13	Schedule It to Run Every Day
14	What's Next? Keep Exploring
15	Make Your Workflow Smarter with Conditional Logic
16	Connect a Live Data Source to Your Workflow

---

© GitHub • Code of Conduct • MIT License