Run Steel on MIOSA
~15 min TypeScript PythonWhat you’re building: Run open-source browser infrastructure for your agents.
Primitives you’ll use: Computer
Agent prompt
Start with your coding agent
Choose how you will use it. The brief installs and runs the real tool on MIOSA, then shapes it for your team or your customers. Copy it into OSA, Claude Code, Codex, or Cursor.
Template coming soon1 What are you building?
Leave it empty and your agent will propose 3 names, pick one, and use it for resources, the domain and branding.
3 Configure
Reference tool
Steel, by Steel (https://steel.dev): open-source browser infrastructure that lets agents control fleets of cloud browsers.
How it works: Steel exposes a browser API: you create sessions, drive them with an automation library, and read back pages, screenshots and PDFs.
Key capabilities:
Cloud browser sessions through an API
Controlled by Playwright, Puppeteer or an agent
Screenshots and page extraction
Open source
You will install and RUN the real Steel (open source browser API). Read its docs first: https://github.com/steel-dev/steel-browser. Every install command below comes from them; if the docs and this brief disagree, the docs win.
How it runs on MIOSA
Where it runs -> a MIOSA computer (a persistent desktop with a browser)
State and files -> Run it as a background process so it survives your shell; session data is transient by design.
Model -> your own provider key, never a MIOSA platform key
Its web port 3000 -> a MIOSA preview URL (`miosa preview create`)
Your customers -> one workspace each, isolated from each other
Goal
Install and run Steel on a MIOSA computer, so it can serve my customers. Each customer is isolated in its own workspace; I meter usage and bill them.
Product name: propose 3 product names, pick one, and use it for resource names, the domain and branding. Until you pick, <product-name> stands for it in the commands below.
Scale: a prototype.
Set up
npm i -g @miosa/cli
miosa login && miosa whoami
miosa api-key create <product-name>-key --preset agent
export MIOSA_API_KEY="msk_u_..."
miosa org
Resources
Computer: a persistent Linux desktop with a browser
miosa computer create <product-name>-desktop
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 <product-name>-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).
Install Steel
Run these inside the computer (its terminal), in order. The commands are from the official docs.
git clone https://github.com/steel-dev/steel-browser.git
cd steel-browser
npm install
Check: the install finishes without errors; the next sections configure and start it.
Configure
Steel needs Chrome: on a MIOSA computer set CHROME_EXECUTABLE_PATH to the installed Chrome. The README also documents a Docker image if you prefer it.
Model: chosen by your agent. Use a current model from a provider I have connected; ask which if none is.
Run it
Start it and prove it works.
npm run dev
curl -X POST http://localhost:3000/v1/sessions # create a session
Check: the tool starts without errors and responds.
Persistence and access
Run it as a background process so it survives your shell; session data is transient by design.
It listens on port 3000. Run it as a background process so it outlives your shell, then open a preview:
miosa preview create <product-name>-desktop 3000 --name web
Check: the preview URL answers, and still answers after the machine pauses and resumes (previews wake it on request).
Make it a product
One Steel computer per customer (or a pool with a session router) behind your own API keys; never expose the raw port publicly.
Limits and costs
Pin the version you tested and update deliberately; these tools change quickly.
Give the tool only the credentials it needs, as secrets, not baked into files.
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
Steel runs on a MIOSA computer and answers a real task.
State survives a pause and resume.
Each customer has its own workspace and its own instance; nothing is shared between them.
Everything it created can be deleted with nothing left running.
What your choices added
- Build a product. a workspace per customer, per-customer metering and bill-to
- Agent suggests a name. proposes 3 product names and picks one; commands use <product-name>
- Model: chosen by your agent. your agent picks a model
- Prototype. one small machine, no extras, easy to delete
What you're building
open-source browser infrastructure that lets agents control fleets of cloud browsers Modelled on Steel, by Steel.
Steel exposes a browser API: you create sessions, drive them with an automation library, and read back pages, screenshots and PDFs.
Primitives you'll use: Computer
- Cloud browser sessions through an API
- Controlled by Playwright, Puppeteer or an agent
- Screenshots and page extraction
- Open source
You install and run the real Steel on MIOSA. Read the official docs first: github.com/steel-dev/steel-browser. Every install command below comes from them; if the docs and this guide disagree, the docs win. Steel is open source browser API.
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 agentDocs - Computer A persistent Linux desktop for browser, mouse, and screenshot work.
miosa computer create app-desktopDocs - A workspace per customer Isolates each customer’s machines, deployments, and data as they onboard.
miosa workspace create customer-1Docs - Branding and white-label Your name and slug on previews, deployments, and the desktop; customers never see MIOSA.
miosa orgDocs - Usage metering and bill-to Usage per customer, and which account pays for new machines.
miosa org billDocs
Architecture
How it runs on MIOSA
Where it runs -> a MIOSA computer (a persistent desktop with a browser)
State and files -> Run it as a background process so it survives your shell; session data is transient by design.
Model -> your own provider key, never a MIOSA platform key
Its web port 3000 -> a MIOSA preview URL (`miosa preview create`)
Your customers -> one workspace each, isolated from each other
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 steel-customer-1Tag 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).
Install Steel
Run these inside the computer (its terminal), in order. The commands are from the official docs.
git clone https://github.com/steel-dev/steel-browser.gitcd steel-browsernpm installCheck: the install finishes without errors; the next sections configure and start it.
Configure
Steel needs Chrome: on a MIOSA computer set CHROME_EXECUTABLE_PATH to the installed Chrome. The README also documents a Docker image if you prefer it.
Model: chosen by your agent. Use a current model from a provider I have connected; ask which if none is.
Run it
Start it and prove it works.
npm run devcurl -X POST http://localhost:3000/v1/sessions # create a sessionCheck: the tool starts without errors and responds.
Persistence and access
Run it as a background process so it survives your shell; session data is transient by design.
It listens on port 3000. Run it as a background process so it outlives your shell, then open a preview:
miosa preview create steel-desktop 3000 --name webCheck: the preview URL answers, and still answers after the machine pauses and resumes (previews wake it on request).
Make it a product
One Steel computer per customer (or a pool with a session router) behind your own API keys; never expose the raw port publicly.
Limits and costs
Pin the version you tested and update deliberately; these tools change quickly.
Give the tool only the credentials it needs, as secrets, not baked into files.
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
Steel runs on a MIOSA computer and answers a real task.
State survives a pause and resume.
Each customer has its own workspace and its own instance; nothing is shared between them.
Everything it created can be deleted with nothing left running.