Agent Control Walk-Through: Open-Source Control Plane for AI Agents

In this walkthrough, we explore Agent Control, the open-source control plane for AI agents, using a live multi-agent stock-scoring system. Watch as two agents debate market data, leak sensitive client information, and see the issue resolved in under two minutes with zero downtime. No code patch. No redeployment. Just a policy toggle in the Agent Control UI. You'll see: → How the @control() decorator and Agent Control server work end-to-end → A live PII detection and redaction fix applied at runtime → The real-time audit trail and fleet-wide dashboard → How to create custom controls and apply them across any agent, project, or team ⭐ Star the repo: https://github.com/agentcontrol/agent-control 📖 Docs: https://agentcontrol.dev/ 📅 Register for the webinar: https://galileo.ai/webinar/agent-control-centralized-governance-for-every-agent 📝 Read the launch blog: https://galileo.ai/blog/announcing-agent-control Agent Control is open source under Apache 2.0, vendor-neutral, and works with the frameworks and guardrail providers you already use. About Galileo Galileo is the agent observability and guardrails company that helps enterprises run trustworthy agents. Learn more at https://galileo.ai Chapters 0:00 — The problem with hard-coded guardrails 0:24 — Introducing Agent Control 0:39 — Architecture and integration 1:03 — Live demo: multi-agent stock scoring system 1:24 — Catching a PII compliance incident 1:30 — Fixing it in the Agent Control UI — no redeployment 2:06 — Real-time audit trail and dashboard 2:33 — Custom controls: build once, apply everywhere 2:54 — Get started