Authenticating AI Agents

As AI agents continue to grow in popularity, figuring out how to keep them from leaking secrets, destroying sensitive data, or otherwise wreaking havoc is more important than ever. Chris Sev teaches us how to control what agents can & can’t access. ---------------------------------------------------------------------- CodeTV supporters get early access to new videos, exclusive bonus content, and other perks. Join us in building a happier, healthier, friendlier web dev community! https://codetv.dev/support ---------------------------------------------------------------------- Chapters: 00:00 Welcome and the big AI risk 00:44 Meet Chris Sev at Auth0 01:45 Auth0 AI agent pillars 04:53 Clone and tour the demo app 06:46 Run the chatbot locally 09:31 Add Auth0 login to Next.js 14:22 Configure env and callbacks 20:58 Fix middleware and cookie error 24:53 Login flow and success 26:47 Next up Token Vault setup 28:06 Enable Token Vault APIs 29:09 Create GitHub Connection 31:07 GitHub OAuth App Setup 35:35 Add Secrets to Auth0 37:38 Turn On Token Vault Grant 38:32 Code Auth0 AI Client 41:43 Build List Repos Tool 45:56 Wire Tools Into Chat Route 48:11 Debug Hanging Tool Calls 50:40 Switch to Working Branch 52:39 Logs and Audit Trail 57:26 Demo App and Wrap Up