Vibe Check: Secrets Management for Agentic Systems

​AI agents are great at calling tools and APIs — but every one of those calls needs credentials, and that's an attack surface most agent demos never talk about. Hand an agent a database password or an API key directly, and you've created a security problem that's easy to miss until something goes wrong.

​In this VibeCheck, Cecil will be chatting with up with Jake from Infisical, a security platform for managing secrets, certificates, and access across your infrastructure (including your AI agents), to talk through how to actually think about this problem.

​We'll have a live demo of Infisical's Agent Proxy, which lets agents use tools and APIs without ever exposing the underlying credentials to the agent itself. Also, we'll get into why agents need security in the first place, where this shows up in real systems people are building today, and how Infisical helps us build secure applications .

​What you'll walk away with:

• ​A live look at Infisical's Agent Proxy

• ​A clear picture of the security risks unique to agentic systems

• ​Real-world use cases where credential exposure is a genuine risk

• ​Practical steps for securing the agents you're building or operating