Build an internal AI workspace

~25 min TypeScript Python

What you’re building: One place for your team’s agents, apps, and data.

Primitives you’ll use: Sandbox, Agent and harness, Deployment (Docker Host)

Agent prompt

Start with your coding agent

A few quick questions, then copy a prompt for any coding agent.

Who is it for?

Pick one to start. The rest adapts to it.

For my company Internal. One organization, a workspace per team, SSO. No billing. Build a product For customers. A workspace per customer, isolation, metering and billing.

What you're building

the one place a team goes to ask, build and ship with agents

Everyone is a member of a single organization; each team has its own workspace with its machines, deployments and data, and agents act with the team's permissions.

Primitives you'll use: Sandbox · Agent and harness · Deployment (App Engine)

Chat with agents that do company work
A workspace per team
Agents that build and run internal tools
Access control and an audit log

This is a pattern, not a copy of one product. Start from the related MIOSA guide and build the smallest version that does the capabilities above.

What you need on MIOSA

Each row is one thing to create before you start. The number matches the step that uses it.

Organization and API key Scopes every call; a workspace key is all a worker needs. miosa api-key create app-key --preset agent Docs
Sandbox The isolated Linux workspace the agent writes code and runs commands in. miosa create app-box --template nextjs --wait Docs
A workspace per customer Isolates each customer’s machines, deployments, and data as they onboard. miosa workspace create customer-1 Docs
Branding and white-label Your name and slug on previews, deployments, and the desktop; customers never see MIOSA. miosa org Docs
Usage metering and bill-to Usage per customer, and which account pays for new machines. miosa org bill Docs
Agent and harness Turns a prompt into work: pick the harness and model a run uses. miosa agent harnesses Docs
Deployment (App Engine) Publishes the app to an immutable, versioned URL with rollback. miosa deploy create --from-sandbox app-box --name app --wait Docs

Architecture

How it maps onto MIOSA

One organization for the company -> the MIOSA organization, with teammates as members (`miosa member add`)

A workspace per team -> `miosa workspace create`

Agents that do work -> agent runs (`miosa prompt`, client.runs) streaming events in a sandbox

Audit -> `miosa audit`

Data and storage

Keep application data in managed Postgres with a row-level owner (member id or team), and keep an append-only audit table for agent actions.

Auth and tenancy

Your organization is the platform; customers never get a MIOSA account, they sign into YOUR product. One workspace per customer, created as they onboard, isolates their machines, runs and data.

miosa workspace create team-ai-workspace-customer-1

Tag every machine and run with the customer id in `metadata`, and never query across customers. Meter usage per customer with GET /api/v1/usage and set who pays with PUT /api/v1/bill-to (/docs/platform/usage-and-billing).

Agent loop

Harness: OSA is MIOSA's own harness and works with any model provider you connect, including your own model.

Model: Anthropic (Claude). Calls use my own provider key (`miosa connections add models`); MIOSA platform keys are never used.

Sessions: one chat per project or conversation, so the agent keeps its context. The first `miosa prompt` on a sandbox uses `--new-chat` (a chat id is printed); every later turn passes `--chat <chat-id>`. `--reuse chat` keeps one new machine per chat so files persist.

Streaming: follow a run with `miosa run follow <run-id>` or client.runs.streamEvents(run.id), and steer or stop it with `miosa run steer` and `miosa run interrupt`.

Agents act as members: every run carries the asking member's id in `metadata`, uses credentials scoped to that member or team, and cannot exceed what the member may do.

Steps

Create a workspace per team inside the one organization, and invite members.
```
miosa workspace create team-ai-workspace-ops
```
```
miosa org invite dana@team-ai-workspace.com
```
Check: A member of one team cannot see another team's machines.
Run the chat backend and agent runner in a sandbox in each team workspace.
```
miosa create team-ai-workspace-box --template agent-node --wait
```
Check: Each team has its own backend.
Dispatch agent work from the chat, on the team's sandbox, with the member's id in run metadata.
```
miosa prompt --sandbox team-ai-workspace-box --harness osa --model <anthropic-model-id> --chat <chat-id> "<request from the member>"
```
Check: Every run shows who asked and which team it ran for.
Control access: members and roles, and an audit trail of what agents did.
```
miosa member add dana@team-ai-workspace.com --role member
```
```
miosa audit
```
Check: A removed member loses access immediately.

Publish the internal UI.

miosa deploy create --from-sandbox team-ai-workspace-box --name team-ai-workspace --dir /workspace --port 3000 --run-command "npm start" --wait

Check: Members can open it and nobody else can.

Limits and costs

Connectors are the real work: budget time for each source system's auth, rate limits and data shape.

Set `idle_timeout_sec` so unused team sandboxes pause.

Prefer read-only access by default; grant writes per task.

Prototype: keep it to one machine at the default size, skip replicas and custom hostnames you do not need, and delete everything when you are done.

Acceptance checks

Members see only their team's data and tools.

Every agent action is attributable to a member in the audit log.

Nothing consequential happens without the approval the guide describes.

Each customer is isolated in its own workspace and usage is metered against them.

Everything it created can be deleted with nothing left running.