i deployed an agent and i got few people to test it and after some time the agent just shut off. it goes from "running" status to "pending" in the dashboard. i don't know why and it's the second time its happening to me

Thanks for reaching out, Farel. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

agent id is : "CA_uBPk3h4d75aC" and : lk agent logs --id CA_uBPk3h4d75aC WARNING: config file C:\Users\User/.livekit/cli-config.yaml should have permissions 600 WARNING: config file C:\Users\User/.livekit/cli-config.yaml should have permissions 600 Using project [ayah-voice-medical-profile] Using agent [CA_uBPk3h4d75aC] failed to get logs: The agent has shut down due to inactivity. It will automatically start again when a new session begins.

Thanks — I looked at the logs you pasted for agent

CA_uBPk3h4d75aC

. What happened • The agent shut down due to inactivity. The CLI message “The agent has shut down due to inactivity. It will automatically start again when a new session begins.” means the agent was scaled down (sleeping), which can show as “pending”/waking in the dashboard while it restarts. What you can do next • If this behavior is acceptable: no action is required — the agent will automatically restart when a new session begins (expect a cold-start delay ~10–20s). • If you want the agent to stay running continuously: ◦ Move to a plan that keeps active replicas (paid tier) or ◦ Keep the agent “warm” by sending small periodic traffic/heartbeats that create an active session. If you want, tell me: • Which plan you’re on (Build/Team/etc.) and • Whether you prefer preventing sleeps or tolerating cold starts I’ll recommend the exact next steps based on your plan and goal. Sources: Deploying to production | LiveKit Docs | Quotas and limits | LiveKit docs | Agent deployment CLI reference | LiveKit docs

i am on the free plan

Thanks — since you’re on the free plan, that explains it. What’s happening • Free-plan agents are automatically scaled down after inactivity. The dashboard shows this as “pending”/waking while the agent restarts. The CLI log message you saw confirms the agent shut down due to inactivity. What you can do • Accept the behavior: agent will auto-restart when a new session begins (expect a ~10–20s cold-start delay). • Prevent sleeps: ◦ Upgrade to a paid plan that keeps replicas active, or ◦ Implement a keep-alive that periodically creates activity (a lightweight session/heartbeat) so the agent stays warm. Which would you prefer — help with upgrading steps, or a short walkthrough to add a keep-alive?