Build an OpenSwarm-style multi-agent app

What you’re building: a multi-agent system: one request is split into parallel tasks, each handled by its own agent in its own machine, and the results are joined.

Primitives you’ll use: Organization, Sandboxes, Agents, Deployments, Custom domains

What you’re building

Multi-agent products trade a single long run for several short ones. An orchestrator decomposes the request, dispatches a child agent per subtask into its own machine, waits for them, and joins the results.

MIOSA is built for this: each child gets its own sandbox so there is no shared state and no interference, and the orchestrator can stream a merged view of every child’s events. Snapshots make fan-out safe (branch from a common starting point), and artifact collection gives the orchestrator the files each child produced.

The pattern is the same one behind agent-team products: an orchestrator, child devices, shared artifacts, and one event stream.

What you need on MIOSA

Architecture

Step 1: Create an account and an API key

Install the CLI, sign in, and mint a key for your product. The secret is printed once; store it as an environment variable, never in code.

npm i -g @miosa/cli
miosa login
miosa whoami
miosa api-key create swarm-app-key --preset agent     # prints the secret once
export MIOSA_API_KEY="msk_u_..."

Both SDKs read MIOSA_API_KEY from the environment, so a server or a worker needs nothing else. See API keys for scopes and how to create keys in the console.

Step 2: Create the resources

Step 3: Wire the agent loop

For a fan-out that starts from the same prepared state, snapshot the orchestrator’s workspace once and restore it into each child.

Step 4: Preview

A Preview is a throwaway public URL for a port inside the machine. Start the app as a background process so it survives the CLI disconnecting, then expose its port.

Open the exact Preview URL MIOSA returns.

Step 5: Deploy

Publish the machine to an immutable Deployment. Give it the source machine, the path the app lives at inside it, and the command that serves it.

Publishing the same deployment name again creates a new immutable version and keeps the same URL. Roll back with miosa deploy rollback swarm-app --to <version-id>, and inspect a live app with miosa deploy logs swarm-app -n 100. See Publishing and Rollback.

Step 6: Custom domain

Attach a domain your customer owns to the deployment. The platform URL keeps working while DNS propagates.

miosa deploy domain-add swarm-app app.example.com
miosa deploy domains swarm-app          # shows the DNS record and verification target
miosa deploy domain-verify swarm-app <domain-id>

Copy the exact DNS record MIOSA shows into the customer’s DNS provider; for a subdomain it is normally a CNAME. Once verified and TLS is active, MIOSA routes the domain to the deployment’s active version. In code, read the URL from the publish response instead of building it: the SDK method is miosa.deployments.domains.add(deploymentId, { domain }) (TypeScript) or miosa.deployments.domains(deployment_id).add(domain=...) (Python). See Domains.

Costs and limits

Everything bills while it runs: machines (sandboxes and computers) by the second, managed databases while running, and deployments when they serve. Pausing a machine stops compute billing but keeps disk. Set an --idle-timeout or a --ttl so a forgotten machine cannot run forever, and cap spend with rate limits and per-workspace quotas.

• Pricing - plans, the published rate card, and the free Developer grant.
• Sizing and limits - machine sizes and what each costs to run.
• Timeouts and auto-stop - --ttl, --idle-timeout, and pause/resume.

Next steps