agentctl: A CLI for Managing AI Agents & MCP Servers Like Infra
summary
A Go-based CLI called agentctl manages hundreds of AI agents across ~50 environments and MCPs, providing auto-discovery of agents, a real-time dashboard of tool calls (Kasa MCPs, llama visual studio), health checks, logs and telemetry, and exposes SLO/SLI metrics to Grafana. It captures token usage, detects hallucinations with confidence-gating alerts, surfaces reasoning chains for audits, and supports direct commands to agents; the tool is open source and reported to run in production at scale.
AI-written summary · about
chapters (12)
- 0:00Agent Ops Challenge
- 0:27Introducing Agent CTL
- 1:03Unified Visibility Needs
- 1:27Discovery and Commands
- 2:19Agent Observability Analogy
- 2:37Guardrails and SRE Metrics
- 3:31Dashboards and Monitoring
- 3:49CLI Walkthrough
- 4:44Health Checks and Logs
- 5:44Reasoning Trust and Alerts
- 7:03Integrations and Wrap Up
- 7:35Closing Thanks
related talks
40mAgents in Action: Advancing Open Source AI Missions
36mPutting AI Agents to Work: A New Era of Open Source Connectivity with MCP
01h 28mAI Coding Agents and how to code them
40mBuilding Agents with Agent Bricks and MCP
46mMCP and A2A Revolutionizing Interconnection and Collaboration of AI Agents
22mThe Agent Control Plane: Managing Thousands of AI Agents
45mAgents, Tools, and MCP, oh my! Next-level AI concepts for developers
18mThe "Agentic Organisation": Governing Multi-Agent Developer Teams with Nutanix Agent Gateway and MCP