Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
Shai-Hulud: Agentic Secret Protection
Learn how to build an agent that patches SCA vulnerabilities and protects code from malicious package secret exfiltration, running securely without static credentials.
I built an agent that patches SCA vulnerabilities and protects a code base against malicious packages that exfil secrets, a common security incident known as Shai Halud. The agent itself is protected from the effects of Shai Halud since it runs without a .env.
Keycard mints short-lived, task-scoped, least-privilege credentials for AI agents.