ASAgentStead Docs

How AgentStead Works

A plain-language view of the AgentStead platform — what happens when you deploy, and what the platform manages for you.

AgentStead is a managed platform for deploying isolated agent environments on runtime hosts. From your perspective it feels simple. Under the hood it's coordinating subscription capacity, provisioning, credentials, backup storage, and runtime access across a fleet.

What happens when you deploy

Deployment typically completes in 30–60 seconds. The status stream in the dashboard updates in real time.

Regions

Agents deploy into one of two regions. Region affects latency and runtime placement — not the dashboard workflow.

What AgentStead manages for you

Runtime host selection

AgentStead picks the right node in your chosen region.

Provisioning & restarts

Container lifecycle is fully managed — no Docker commands from you.

Public routing & TLS

Every agent gets a dedicated HTTPS URL, automatically.

Credential bootstrap

Username, password, and runtime token generated and stored securely.

Backup accounting

Quota tracking and backup catalog across all your agents.

Health visibility

Status, logs, and restart controls all exposed in the dashboard.

What you manage

Secrets & API keys

Add your provider keys and env vars in the Secrets tab.

Runtime configuration

Model settings and structured config in the Settings tab.

Workspace contents

What's inside the persistent workspace is yours to manage.

Backups

Create backups before risky changes — AgentStead doesn't auto-backup.

Active agents

Running agents consume a slot. Destroy what you don't need.