# Cua Documentation

> Cua is an open-source platform for computer-use automation: AI agents that operate real computers by clicking, typing, and reading the screen and accessibility tree. Drive a machine you already have with Cua Driver, provision hosted desktops with Cloud Fleets and the Sandbox SDK, or use Lume as local Apple Silicon/macOS VM infrastructure underneath your own workflows.

Full text of every page: https://cua.ai/docs/llms-full.txt

## Start here

- [Overview](https://cua.ai/docs.md): Computer-use infrastructure on your own machines. Sandboxes, Spaces, and agents that drive real desktops.
- [Sandbox images](https://cua.ai/docs/start-here/sandbox-images.md): The desktop and benchmark images Cua publishes.
- [Create a Space with the Cua SDK](https://cua.ai/docs/start-here/create-a-space-with-the-sdk.md): Create a Linux Space on your own machine from Python or TypeScript, run a command in it, take a screenshot, ask Cua Driver for its windows, watch it, and delete it.
- [Host Spaces on your spare Mac](https://cua.ai/docs/start-here/host-spaces-on-your-spare-mac.md): Turn a spare Mac mini (or any machine) into a place your laptop and your agents create Spaces on, over the relay, ssh or Tailscale, without sharing its own desktop.
- [Drive your first app with Cua Driver](https://cua.ai/docs/start-here/drive-your-first-app.md): Install Cua Driver, connect your agent, and verify a result in a desktop app on macOS, Windows, or Linux.
- [Create a local macOS sandbox with the Cua CLI](https://cua.ai/docs/start-here/create-a-local-macos-sandbox.md): Create a macOS Tahoe sandbox on an Apple silicon Mac with `cua sb create macos`, run a command, take a screenshot, open its desktop, and remove it.
- [Run a sandbox in GitHub Actions](https://cua.ai/docs/start-here/run-in-github-actions.md): Start a Linux sandbox in a GitHub Actions job with the cua-sandbox action, run a command in it, keep a screenshot, and delete it.
- [Build your first Cua Bench task](https://cua.ai/docs/start-here/your-first-cua-bench-task.md): Create a Cua Bench task, score it with its oracle in a local Linux sandbox, solve it yourself, then hand it to an agent.

## Cua SDK

- [Overview](https://cua.ai/docs/cua-sdk.md): Build agentic applications and workflows on isolated Linux, Windows and macOS desktops on your machine, or on any machine that runs cua-spacesd.
- [Quickstart](https://cua.ai/docs/cua-sdk/quickstart.md): Your first sandbox. Run a web server in a sandbox on your machine and call it.

### Guides

- [Choose and build an image](https://cua.ai/docs/cua-sdk/guides/images.md): Pick a canonical or registry image, add packages and files, and pull private images, the same way locally and in the cloud.
- [Run and share a service](https://cua.ai/docs/cua-sdk/guides/services.md): Run a server in a sandbox, call it, connect to an MCP server in it, forward a port, and share it with a public URL.
- [Pass secrets into a sandbox](https://cua.ai/docs/cua-sdk/guides/secrets.md): Pass credentials from the host into a local or cloud sandbox at runtime, and keep them out of images.
- [Add sidecar containers](https://cua.ai/docs/cua-sdk/guides/sidecars.md): Run extra containers such as a database or cache next to a sandbox and reach them by name.
- [Manage sandbox lifecycle](https://cua.ai/docs/cua-sdk/guides/lifecycle.md): Create, keep, reconnect to, list, suspend and delete sandboxes, keep cloud sandboxes alive, and run many at once.
- [Control the desktop](https://cua.ai/docs/cua-sdk/guides/desktop.md): Take screenshots, click and type, call typed Cua Driver tools, open a shell or a browser view, and attach to any cua-spacesd.
- [Space thumbnails](https://cua.ai/docs/cua-sdk/guides/thumbnails.md): Read a Space's latest small preview from the cache every client on the machine shares.
- [Attach to or run cua-spacesd](https://cua.ai/docs/cua-sdk/guides/spacesd.md): Use any machine that runs cua-spacesd as a sandbox, run and configure cua-spacesd yourself, and publish it through the relay.
- [Set up local runtimes](https://cua.ai/docs/cua-sdk/guides/local-runtimes.md): Which runtime runs a local sandbox, how to check and set up the host, and how to pick a kind or runtime.
- [Disk usage and cleanup](https://cua.ai/docs/cua-sdk/guides/disk-usage.md): What local sandboxes keep on disk and how the SDK keeps it bounded.
- [Run sandboxes in the cloud](https://cua.ai/docs/cua-sdk/guides/cloud.md): Run the same sandbox code in the Cua cloud, make the cloud your default, set up credentials, and see what differs from local.
- [Use your own cloud](https://cua.ai/docs/cua-sdk/guides/your-cloud.md): Run sandboxes and Spaces in your own AWS, Google Cloud or Modal account with --on aws, --on gcp or --on modal, with Cua's tags, time limits and cleanup on everything it creates.
- [Your cloud resources and costs](https://cua.ai/docs/cua-sdk/guides/your-cloud-resources.md): What Cua creates in your AWS, Google Cloud or Modal account, what it costs per hour, and the notes for each cloud.
- [Run on other sandbox providers (contrib)](https://cua.ai/docs/cua-sdk/guides/contrib-providers.md): Run cua sandboxes on E2B, Daytona or Modal with --on e2b, --on daytona or --on modal, from the same registry images, with cua-spacesd, services and cua doctor.
- [Test in CI with Cua sandboxes](https://cua.ai/docs/cua-sdk/guides/test-in-ci.md): Run tests against a Cua sandbox in GitHub Actions or any CI, inject the build under test, and fail on stale builds.
- [Run a coding agent in a sandbox](https://cua.ai/docs/cua-sdk/guides/run-a-coding-agent.md): Start Claude Code, Codex, Gemini CLI and other harnesses inside a sandbox, stream their events, follow up, interrupt, and collect results and files.
- [Configure coding agents](https://cua.ai/docs/cua-sdk/guides/coding-agent-options.md): Give a sandboxed coding agent MCP servers, a custom endpoint, and credentials; the supported harnesses and the event kinds they emit.
- [Browse the web from your agent with Cua](https://cua.ai/docs/cua-sdk/guides/browse-the-web.md): Give a coding agent (Claude Code, Codex, Cursor and others) a real browser inside a disposable sandbox with nothing but the cua MCP server. Discover an image, create a browser sandbox, read, click, fill and screenshot pages, teleport a signed-in session with consent, and clean up.
- [Run agent frameworks in a sandbox](https://cua.ai/docs/cua-sdk/guides/agent-frameworks.md): Run OpenAI Agents, Codex, Cursor, OpenClaw, an Omarchy desktop and a Minecraft agent in Cua sandboxes.
- [Omarchy with Cua Driver](https://cua.ai/docs/cua-sdk/guides/omarchy.md): Run the Omarchy edge VM image, where Omarchy's own Cua Driver injects input at the OS level, and drive it with the cua CLI.
- [Migrate from deprecated packages](https://cua.ai/docs/cua-sdk/guides/migrate-from-deprecated-packages.md): cuabot, cua-computer, cua-computer-server and cua-agent are no longer maintained. What each one did and what to use instead.

### Concepts

- [How sandboxes work](https://cua.ai/docs/cua-sdk/concepts/how-sandboxes-work.md): The SDK, the daemon, cua-spacesd and the backends, where a sandbox runs and on what, how it becomes ready, and how its lifetime works.
- [Telemetry and privacy](https://cua.ai/docs/cua-sdk/concepts/telemetry.md): What the cua SDK, the cua CLI, cua daemon, cua-spacesd and the Cua Spaces app collect, what they never collect, and how to turn it off.
- [Telemetry events](https://cua.ai/docs/cua-sdk/concepts/telemetry-events.md): Every usage telemetry event the Cua programs send, each property it carries, and why.

### Reference

- [Reference](https://cua.ai/docs/cua-sdk/reference.md): The Cua SDK reference: every object with its signature in Python, TypeScript, Swift, Kotlin and Rust.

## Sandboxes

- [Cua client](https://cua.ai/docs/cua-sdk/reference/cua.md): The Cua handle: embedded or daemon client, its configuration, and the SDK version.

#### Sandbox

- [Sandbox](https://cua.ai/docs/cua-sdk/reference/sandbox.md): Create, find, connect to and control sandboxes, locally, in the cloud or by address.
- [Lifecycle](https://cua.ai/docs/cua-sdk/reference/sandbox/lifecycle.md): Wait for, keep alive, suspend, resume, restart, detach and delete a sandbox.
- [Without cua-spacesd](https://cua.ai/docs/cua-sdk/reference/sandbox/without-spacesd.md): Run commands, take screenshots and open the display of a local Lume sandbox that has no cua-spacesd.
- [Create options](https://cua.ai/docs/cua-sdk/reference/sandbox/options.md): SandboxCreateOptions and what it holds: cloud options, readiness probes and sidecar containers.
- [Image](https://cua.ai/docs/cua-sdk/reference/image.md): Canonical images, image references, build layers and registry credentials.
- [Image catalog details](https://cua.ai/docs/cua-sdk/reference/os-image-catalog.md): Every image variant, its runtimes, guest services and limits.
- [What's installed](https://cua.ai/docs/cua-sdk/reference/image-software.md): The apps, tools and simulator runtimes each image tier ships, as the strict doctor measured them.
- [Services](https://cua.ai/docs/cua-sdk/reference/services.md): Named services, port forwards, public URLs and MCP clients of a sandbox.
- [Local runtimes](https://cua.ai/docs/cua-sdk/reference/local.md): The container and VM runtimes this machine can host, their setup, and local images.
- [Runtime support](https://cua.ai/docs/cua-sdk/reference/runtime-support.md): Which images run on which local runtime or in the cloud, what customization each accepts, and how ports reach them.

## Cloud capacity

#### Fleet

- [Fleet](https://cua.ai/docs/cua-sdk/reference/fleet.md): Cloud pools and templates: apply, check, export, scale and delete (advanced).
- [Claims](https://cua.ai/docs/cua-sdk/reference/fleet/claims.md): Claim sandboxes from a cloud pool, keep them alive, release them, and sign service URLs.
- [Cloud images and runtimes](https://cua.ai/docs/cua-sdk/reference/fleet/images.md): Fleet image resources and the runtime (gVisor or KubeVirt) each image runs on.

## Inside a sandbox

#### Spacesd client

- [Spacesd client](https://cua.ai/docs/cua-sdk/reference/spacesd.md): SpacesdClient: the cua-spacesd client for processes, files, the desktop and media in a sandbox.
- [Processes](https://cua.ai/docs/cua-sdk/reference/spacesd/shell.md): Run, spawn and attach to processes in a sandbox through cua-spacesd.
- [Files](https://cua.ai/docs/cua-sdk/reference/spacesd/files.md): Upload, download, list and remove files in a sandbox through cua-spacesd.
- [Computer](https://cua.ai/docs/cua-sdk/reference/spacesd/computer.md): Screenshots, pointer, keyboard, clipboard and displays of a sandbox desktop through cua-spacesd.
- [Media](https://cua.ai/docs/cua-sdk/reference/spacesd/media.md): Video and audio streaming sessions and the sink callbacks you implement.
- [Coding agents](https://cua.ai/docs/cua-sdk/reference/agents.md): Run coding agents (Claude Code, Codex, Gemini CLI, OpenCode, ...) inside any sandbox.

## Spaces

#### Spaces

- [Spaces](https://cua.ai/docs/cua-sdk/reference/spaces.md): The Spaces registry: create, add, resolve, delete and remove Spaces.
- [Create options](https://cua.ai/docs/cua-sdk/reference/spaces/create.md): What `Spaces.create` takes and reports: options, progress, cancel and GPU acceleration.
- [Space](https://cua.ai/docs/cua-sdk/reference/spaces/space.md): A connected Space: shell, files, tools, windows and the hotspot.
- [Cua Volume](https://cua.ai/docs/cua-sdk/reference/spaces/volume.md): The versioned volume every Space and agent shares: files, history, grants, requests and audit.
- [Cua Volume storage and sync](https://cua.ai/docs/cua-sdk/reference/spaces/volume-sync.md): Where the drive keeps its bytes, the volume on this machine and in Spaces, sync across devices, and the block cache.
- [Streams and presence](https://cua.ai/docs/cua-sdk/reference/spaces/streams.md): Media sessions of a Space and presence (who is connected, and their cursors).
- [Windows, icons and displays](https://cua.ai/docs/cua-sdk/reference/spaces/windows.md): A Space's app icons (batched, through the SDK's one icon cache), window previews, displays and memory and storage use.
- [Agents in a Space](https://cua.ai/docs/cua-sdk/reference/spaces/agents.md): Start, follow, message and stop coding-agent runs inside a Space.
- [Persistent agents](https://cua.ai/docs/cua-sdk/reference/spaces/persistent-agents.md): Named agents whose memory outlives their Space: homes in the Cua Volume, routines, notifications, pause and resume, access to your computers.
- [Your cloud](https://cua.ai/docs/cua-sdk/reference/spaces/cloud.md): Sandboxes and Spaces in your own AWS, Google Cloud or Modal account: connect, check, and sweep what Cua created there.
- [Teleport](https://cua.ai/docs/cua-sdk/reference/spaces/teleport.md): Move a running app between machines, with an approval callback.
- [Decoded media](https://cua.ai/docs/cua-sdk/reference/spaces/decoded-media.md): Media sessions that deliver decoded BGRA frames and PCM audio.
- [Host teleport](https://cua.ai/docs/cua-sdk/reference/spaces/teleport-host.md): The Teleport object of the Spaces apps: providers, manifests and sends from this machine.
- [Teleport an app](https://cua.ai/docs/cua-sdk/reference/spaces/teleport-apps.md): The app catalog, plans with consent, runs, drops and window drags.

##### Spaces app core

- [Spaces app core](https://cua.ai/docs/cua-sdk/reference/spaces/app-core.md): The view models the Cua Spaces apps share: Space list, New Space wizard, teleport review, notch and Keyvault.
- [Space list](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/spaces.md): The Space list (roster), the sidebar, a Space's detail and its remote windows.
- [Creates and deletes](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/creates.md): Spaces being created or deleted: the rows they show at once, their progress and how they hand over to the registry.
- [Stream section and facts](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/stream.md): A Space detail's Stream rows (desktop and windows with their icons and picture in picture) and its facts with live memory and storage use.
- [New Space wizard](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/new-space.md): The New Space wizard: where it runs, image, engine, address and the create call.
- [Image catalog](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/images.md): The images the New Space wizard offers: groups, sizes, the image field and its suggestions.
- [Teleport picker](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/teleport.md): The teleport picker and its catalog.
- [Teleport transfer and drag](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/teleport-transfer.md): The overlays that show a teleport in progress: the transfer overlay and drag-and-drop onto a Space.
- [Teleport plan and run](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/teleport-run.md): The teleport plan, its consent and capabilities, and the run's steps.
- [Teleport review](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/teleport-review.md): The review and consent before a teleport: what to send per site, from the app or from the Keyvault, and the choice remembered for next time.
- [Teleport picker grid](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/teleport-grid.md): The teleport picker's tiles in both apps: tabs, app and window tiles with their icon and preview sources, and the arrow keys.
- [Your cloud](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/cloud.md): The "Connect a cloud" sheet and the New Space wizard's "Your cloud": providers with their found mark, Test that creates nothing, Connect, and what a connected cloud runs at what cost.
- [Window drag trigger](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/drag-trigger.md): When dragging a window opens Teleport to Cua: move versus resize, the notch line and the dwell.
- [Experiments](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/experiments.md): Settings > Experiments: the switches that hide Cua Volume, your own cloud and sharing until they are turned on.
- [Run on](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/run-on.md): The New Space Run on menu: This Mac, your machines that provide Spaces, and connected clouds.
- [Power button](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/power.md): The power button on a Space: suspend and resume, or turn off and on.
- [Notch](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/notch.md): The notch switcher: state, events, layout on the screen and motion.
- [Keyvault](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/keyvault.md): The Keyvault broker client and its overview: items, pending requests, grants, rules and audit.
- [Keyvault screens](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/keyvault-screens.md): The Keyvault page, sidebar, the vault list (apps, sites and items with their locks), the unlock prompt and the approval prompt.
- [Onboarding and settings](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/onboarding-settings.md): First-run onboarding, its card miniatures, the settings file and the hotkey label.
- [About, updates and launch](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/about.md): The About window and its update channel, what the apps do once they launch (the bundled cua install), and the daemon and refresh reports they read.
- [Cua Cloud billing](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/billing.md): The Billing row in Settings and the out-of-credit notice on a refused cloud Space.
- [This machine](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/this-machine.md): The This machine roster entry, its page and the host setup form.
- [Window, menu bar and Settings](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/window.md): The window chrome, a Space's toolbar and sections, the menu bar menu, Settings and the coding agents' rows.
- [Devices](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/devices.md): Settings > Devices: this device's enrollment, the account's devices, Recent Access, and the enroll and approval sheets.
- [Share sheet](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/share.md): The Share sheet: who a Space is shared with (viewer or editor) and the share request to run.
- [Usage telemetry](https://cua.ai/docs/cua-sdk/reference/spaces/app-core/telemetry.md): Which anonymous usage events a Spaces app step means, and sending them.
- [App core: Cua Volume storage, mount and sync](https://cua.ai/docs/cua-sdk/reference/spaces/volume-app-core.md): Settings, Storage (this machine or an S3-compatible bucket, the Finder volume, the cache), the first run's Cua Volume page and the Drive page's devices and conflicts.
- [App core: agents, drive, notifications](https://cua.ai/docs/cua-sdk/reference/spaces/agents-app-core.md): The Agents page (persistent agents, memory, routines, computer access), the Drive page and the notifications the apps post.
- [Devices](https://cua.ai/docs/cua-sdk/reference/spaces/devices.md): This device as a client of the account on the relay: enroll, approve, rename, revoke and the audit log.
- [Host and relay](https://cua.ai/docs/cua-sdk/reference/spaces/host.md): Unattended host setup and the cua-relay directory client.

## Account

- [Auth](https://cua.ai/docs/cua-sdk/reference/auth.md): Sign-in and the credential store the SDK, the CLI and the daemon share.
- [Agent setup](https://cua.ai/docs/cua-sdk/reference/agent-setup.md): Detect coding agents, install cua skills and configure their MCP servers.
- [Telemetry](https://cua.ai/docs/cua-sdk/reference/telemetry.md): Anonymous usage telemetry switches.

## Errors and types

- [Errors](https://cua.ai/docs/cua-sdk/reference/errors.md): Every error the Cua SDK raises: its variant, message, cause and fix.
- [Types](https://cua.ai/docs/cua-sdk/reference/types.md): How the SDK types map to each language, and every object, record, enum and function in one list.

## Language packages

#### Python high-level API

- [Python high-level API](https://cua.ai/docs/cua-sdk/reference/python.md): cua_sandbox and the cua package helpers: packages, configuration, and the page for each class.
- [Sandbox](https://cua.ai/docs/cua-sdk/reference/python/sandbox.md): Sandbox, SandboxInfo, the sandbox() helper and the creation options of cua-sandbox.
- [Image](https://cua.ai/docs/cua-sdk/reference/python/image.md): The Image builder, sidecar containers and registry credentials of cua-sandbox.
- [Pool](https://cua.ai/docs/cua-sdk/reference/python/pool.md): Pools, templates, sandbox specs and pool errors of cua-sandbox.
- [Interfaces](https://cua.ai/docs/cua-sdk/reference/python/interfaces.md): Shell, mouse, keyboard, screen, clipboard, window, terminal, files, apps, mobile and driver interfaces.
- [Services, tunnels and MCP](https://cua.ai/docs/cua-sdk/reference/python/services.md): Named services, signed and public URLs, tunnels and MCP endpoints of a cua-sandbox sandbox.
- [Runtimes](https://cua.ai/docs/cua-sdk/reference/python/runtimes.md): Local runtimes (Docker, QEMU, Lume, Hyper-V, Tart, Android emulator) and local support checks.
- [Configuration and errors](https://cua.ai/docs/cua-sdk/reference/python/configuration.md): configure, login, whoami, transports and the errors cua-sandbox raises.

#### TypeScript additions

- [TypeScript additions](https://cua.ai/docs/cua-sdk/reference/typescript.md): The hand-written helpers of @trycua/cua: embedded, connect, Image, probes, sidecars, specs and MCP.
- [Spaces (TypeScript)](https://cua.ai/docs/cua-sdk/reference/typescript/spaces.md): The @trycua/cua/spaces helpers: threads, events, teleport approvals and typed errors.
- [Spaces transport (TypeScript)](https://cua.ai/docs/cua-sdk/reference/typescript/spaces-transport.md): MCP-over-HTTP and Tauri transports for Spaces in webviews and browsers.
- [Spaces host (TypeScript)](https://cua.ai/docs/cua-sdk/reference/typescript/spaces-host.md): The @trycua/cua/spaces/host helpers for hosting a Space from Node.

#### Rust crates

- [Rust crates](https://cua.ai/docs/cua-sdk/reference/rust.md): The public Rust crates of the Cua SDK, how to depend on them, and what is stable.
- [cua-sdk](https://cua.ai/docs/cua-sdk/reference/rust/cua-sdk.md): The Rust host API: the Cua object and everything the language bindings export, plus the Rust-only items.
- [cua-spacesd-client](https://cua.ai/docs/cua-sdk/reference/rust/cua-spacesd-client.md): The async cua-spacesd client crate (native gRPC or gRPC-Web).
- [cua-fleet](https://cua.ai/docs/cua-sdk/reference/rust/cua-fleet.md): The Fleet (cloud sandboxes) crate: pools, claims, builds and claim secrets.
- [cua-sandbox-core](https://cua.ai/docs/cua-sdk/reference/rust/cua-sandbox-core.md): The daemon-agnostic Sandbox over the Fleet, local and direct providers.
- [cua-spaces: registry and primitives](https://cua.ai/docs/cua-sdk/reference/rust/cua-spaces.md): The Spaces registry and primitives (files, exec, streams, presence, services, teleport, hotspot, relay).
- [cua-spaces: client](https://cua.ai/docs/cua-sdk/reference/rust/cua-spaces-client.md): The Spaces client: transports, model, control and teleport.
- [cua-spaces: agents and MCP](https://cua.ai/docs/cua-sdk/reference/rust/cua-spaces-agents-mcp.md): Agents in Spaces and the Spaces MCP server (stdio and HTTP).

## Protocol

#### Protocol

- [Protocol](https://cua.ai/docs/cua-sdk/reference/protocol.md): The gRPC contracts of cua-spacesd (cua.env.v1) and the cua daemon (cua.daemon.v1): ports, routes, auth and every RPC.
- [System](https://cua.ai/docs/cua-sdk/reference/protocol/env-system.md): cua.env.v1 SystemService: capabilities, init, health, shutdown, viewer tickets and relay attach.
- [Diagnose](https://cua.ai/docs/cua-sdk/reference/protocol/env-diagnose.md): cua.env.v1 Diagnose: the image self-test report `SystemService.Diagnose` streams.
- [Shared types and errors](https://cua.ai/docs/cua-sdk/reference/protocol/env-common.md): The cua.env.v1 types every spacesd service shares: ErrorInfo, principals and geometry.
- [Processes](https://cua.ai/docs/cua-sdk/reference/protocol/env-process.md): cua.env.v1 ProcessService: start, attach to, feed and signal guest processes.
- [Filesystem](https://cua.ai/docs/cua-sdk/reference/protocol/env-filesystem.md): cua.env.v1 FilesystemService: stat, list, watch, read, chunked upload and signed URLs.
- [Computer and driver](https://cua.ai/docs/cua-sdk/reference/protocol/env-computer.md): cua.env.v1 ComputerService and DriverService: screenshots, pointer, keyboard, clipboard, displays and Cua Driver tools.
- [Windows and accessibility](https://cua.ai/docs/cua-sdk/reference/protocol/env-windows.md): cua.env.v1 WindowsService and AccessibilityService: list, watch and arrange windows, launch apps, and the accessibility tree.
- [Streams and presence](https://cua.ai/docs/cua-sdk/reference/protocol/env-stream.md): cua.env.v1 StreamService and PresenceService: media sessions and multiplayer cursors.
- [Teleport, tunnels and the volume](https://cua.ai/docs/cua-sdk/reference/protocol/env-teleport-tunnel.md): cua.env.v1 TeleportService, TunnelService and VolumeService: receiving teleported sessions and files, port forwards, hotspots and the guest mount of Cua Volume.
- [Host Spaces](https://cua.ai/docs/cua-sdk/reference/protocol/env-host-spaces.md): cua.env.v1 HostSpacesService: Spaces a host creates for its owner's devices through the relay.
- [Daemon: sandboxes](https://cua.ai/docs/cua-sdk/reference/protocol/daemon-sandboxes.md): cua.daemon.v1 SandboxService: the local cua daemon API for sandboxes in every location.
- [Daemon: local runtimes](https://cua.ai/docs/cua-sdk/reference/protocol/daemon-runtimes.md): cua.daemon.v1 RuntimeService: the local engines and images the cua daemon orchestrates.
- [Daemon: Spaces](https://cua.ai/docs/cua-sdk/reference/protocol/daemon-spaces.md): cua.daemon.v1 SpaceService: Spaces and the hosts that provide them, through the local cua daemon.
- [Daemon: service](https://cua.ai/docs/cua-sdk/reference/protocol/daemon-service.md): cua.daemon.v1 DaemonService: the local cua daemon identity and the loopback media bridge for webviews.

## Cua Driver

- [Overview](https://cua.ai/docs/cua-driver.md): Drive native apps on macOS, Windows and Linux in the background.
- [Quickstart](https://cua.ai/docs/cua-driver/quickstart.md): Install Cua Driver, connect your agent, and have it compute 6 × 7 in a desktop calculator.

### Guides

- [Connect your agent](https://cua.ai/docs/cua-driver/guides/connect-your-agent.md): Register Cua Driver with Claude Code, Codex, Cursor, Hermes, OpenClaw, local models, and agent SDKs.
- [Use the SDK](https://cua.ai/docs/cua-driver/guides/use-the-sdk.md): Call Cua Driver from Python or TypeScript in process, verify each action, and expose MCP from a signed desktop app.
- [Drive a browser](https://cua.ai/docs/cua-driver/guides/browsers.md): Bind a browser window, snapshot its tab, and navigate, click, and type without raising it; attach to a logged-in Chrome or Edge profile.
- [Parse visual regions](https://cua.ai/docs/cua-driver/guides/visual-regions.md): Parse a screenshot into text and icon regions with the optional perception extension, and let an external chooser such as jev-use pick bounded actions.
- [Record trajectories and history](https://cua.ai/docs/cua-driver/guides/trajectories.md): Record an agent's actions, render them as a zoom-on-click video, and keep an encrypted local Computer History.
- [Set permissions](https://cua.ai/docs/cua-driver/guides/permissions.md): Choose a permission mode, narrow it with a capability manifest, and restrict tools with YAML or Rego policies.
- [Run in a VM or over SSH](https://cua.ai/docs/cua-driver/guides/vms-and-remote.md): Run Cua Driver inside a macOS Lume VM or drive a Windows desktop over SSH.
- [Test cua-driver in a sandbox](https://cua.ai/docs/cua-driver/guides/test-in-a-sandbox.md): Run a cua-driver build against a disposable Linux desktop, locally or in CI, and prove the build under test is the one that runs.
- [Keep Cua Driver running and updated](https://cua.ai/docs/cua-driver/guides/operate.md): Keep the Cua Driver daemon running, update it, choose a release channel, personalize the agent cursor, and control telemetry.
- [Design cursor motions](https://cua.ai/docs/cua-driver/guides/cursor-motion.md): Cua Cursor Motion, the open-source library behind Cua Driver's agent cursor. Plan the six motions, design your own, and play them on a web page.
- [Troubleshoot](https://cua.ai/docs/cua-driver/guides/troubleshoot.md): Fix stale macOS permissions and empty results, and work around known per-app and per-platform limits.
- [Recipes](https://cua.ai/docs/cua-driver/guides/recipes.md): Real desktop workflows with Cua Driver, each with the prompt to give your agent and how to check the result.

### Concepts

- [How Cua Driver works](https://cua.ai/docs/cua-driver/concepts/how-cua-driver-works.md): Observation, targets, the background-first action ladder, verification, and who owns the runtime.
- [Perception extension](https://cua.ai/docs/cua-driver/concepts/perception-extension.md): How the optional cua-perception extension turns one screenshot into regions an agent can act on, and how Cua Driver binds every resulting action to that exact screenshot.
- [Platform support](https://cua.ai/docs/cua-driver/concepts/platform-support.md): Proven operating systems, Linux window systems, and browser routes, with known boundaries and remaining work.

### Reference

- [Reference](https://cua.ai/docs/cua-driver/reference.md): Every cua-driver command and every MCP tool on macOS, Linux and Windows, generated from the driver.

#### CLI

- [CLI](https://cua.ai/docs/cua-driver/reference/cli.md): Cross-platform computer-use automation driver.
- [Tools and MCP](https://cua.ai/docs/cua-driver/reference/cli/mcp.md): Serve MCP, print client config, and list, describe or call tools from the shell.
- [Daemon](https://cua.ai/docs/cua-driver/reference/cli/daemon.md): Run, stop and inspect the daemon, its sessions and autostart.
- [Permissions and diagnostics](https://cua.ai/docs/cua-driver/reference/cli/permissions.md): macOS permission grants, diagnostic probes and install reports.
- [Trajectory recording](https://cua.ai/docs/cua-driver/reference/cli/recording.md): Record, inspect and render action trajectories.
- [Computer History](https://cua.ai/docs/cua-driver/reference/cli/history.md): The encrypted, metadata-only Computer History preview.
- [Configuration](https://cua.ai/docs/cua-driver/reference/cli/config.md): Driver configuration, telemetry, cursor themes and cursor motion.
- [Skills and extensions](https://cua.ai/docs/cua-driver/reference/cli/extensions.md): The agent skill pack, signed extensions and the perception parser.
- [Updates](https://cua.ai/docs/cua-driver/reference/cli/updates.md): Check for and apply updates, and choose the release channel.

#### MCP tools

- [MCP tools](https://cua.ai/docs/cua-driver/reference/mcp-tools.md): Every tool cua-driver serves over MCP on macOS, Linux and Windows, with its parameters.
- [App and window tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/apps-and-windows.md): List, launch, quit, front and arrange apps and windows.
- [Window state tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/window-state.md): Screenshots, accessibility trees, verification and visual regions of a window.
- [Screen tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/screen.md): The desktop, screen size, cursor position and zoomed crops.
- [Click tool](https://cua.ai/docs/cua-driver/reference/mcp-tools/click.md): Click by element or pixel.
- [Double-click and right-click tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/double-and-right-click.md): Double-click and right-click by element or pixel.
- [Pointer tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/pointer.md): Drag, scroll, move the pointer, and press or release buttons.
- [Typing tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/keyboard.md): Type text into a field or window.
- [Key and shortcut tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/keys-and-shortcuts.md): Press keys and keyboard shortcuts.
- [Value and clipboard tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/values-and-clipboard.md): Set element values and read or write the clipboard.
- [Batch tool](https://cua.ai/docs/cua-driver/reference/mcp-tools/batch.md): Run several action tools in one call.
- [Page tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/page.md): Read and act on web pages, and prepare and navigate browsers.
- [Browser input tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/browser-input.md): Exact browser input over CDP: clicks, typing, pointer, dialogs, files and downloads.
- [Session tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/sessions.md): Start, escalate, inspect and end lifecycle sessions.
- [Agent cursor tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/agent-cursor.md): Show, animate and style the agent cursor.
- [Recording tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/recording.md): Record and replay trajectories.
- [Configuration and maintenance tools](https://cua.ai/docs/cua-driver/reference/mcp-tools/maintenance.md): Configuration, permissions, health, updates and extensions.

## Cua Spaces

- [Overview](https://cua.ai/docs/spaces.md): The Cua Spaces app and SDK: register any sandbox or machine once, then stream it, run agents in it and teleport apps into it.
- [Quickstart](https://cua.ai/docs/spaces/quickstart.md): Install the Cua Spaces app and the cua CLI, run a local Space in Docker, and register it from code.

### Guides

- [Add a Space](https://cua.ai/docs/spaces/guides/add-a-space.md): Create a Space on your machine, from any image, follow or cancel the create, add a GPU, add an existing machine, then list, delete or remove it.
- [Set up unattended access](https://cua.ai/docs/spaces/guides/unattended-access.md): Make this machine a Space you reach from your other devices: cua host setup installs cua-spacesd as a service that joins the cua.ai relay, or serves on ip:port.
- [Share a Space](https://cua.ai/docs/spaces/guides/share-a-space.md): Let another cua.ai account watch or use one of your Spaces through the relay, and reach a Space from your other devices.
- [Stream a desktop](https://cua.ai/docs/spaces/guides/stream-a-desktop.md): Stream a Space desktop or one window with audio: open a media ticket, pick a codec and transport, and start from the streaming examples.
- [Send files](https://cua.ai/docs/spaces/guides/send-files.md): Send a host file or folder into a Space: it lands in ~/Downloads, verified by SHA-256.
- [Teleport an app](https://cua.ai/docs/spaces/guides/teleport-an-app.md): Bring an app from this machine into a Space: install it there, open your files in it, or move its signed-in state, with a consent screen for everything that leaves.
- [Persistent agents](https://cua.ai/docs/spaces/guides/persistent-agents.md): A named agent whose memory outlives its runs and its Space: a home in the Cua Volume, routines the daemon fires, notifications, pause and resume, and access to your computers by name.
- [Keyvault](https://cua.ai/docs/spaces/guides/keyvault.md): The Cua Keyvault is a password manager for what teleport moves: cookies, localStorage values, passwords and files, grouped by the app they came from. Every item is locked until you allow unattended access. Search, select several, lock, unlock or delete them, approve requests and turn it off from the Keyvault page.
- [Log in to a site](https://cua.ai/docs/spaces/guides/site-login.md): Let an agent sign in to a website in a Space with a password you saved in the Cua Keyvault. You approve each sign-in; the agent never sees the password.
- [What the relay can see](https://cua.ai/docs/spaces/guides/relay-security.md): A plain accounting of what relay.cua.ai (or any self-hosted cua-relay) can and cannot see, and what we have done and are still doing about it.
- [Cua Volume](https://cua.ai/docs/spaces/guides/cua-volume.md): One versioned drive per account, shared by every Space and agent: agent memory and outputs that outlive any Space, with default access rules, grants you approve, and an audit log.
- [Use Spaces from an agent](https://cua.ai/docs/spaces/guides/use-from-an-agent.md): Give an AI coding agent the Spaces tools through cua daemon mcp, and run agents inside a Space as long-lived threads.
- [Updates](https://cua.ai/docs/spaces/guides/updates.md): How the Cua Spaces app for macOS updates itself, how to pick the Stable or Beta channel, and what it refreshes after an update.

### Examples

- [Cua Bots: persistent bots with their own computers](https://cua.ai/docs/spaces/examples/cua-bots.md): The featured Cua Spaces example: a source-available Mac and iPhone app for persistent bots, each with a face you design, its own Space, a memory in the Cua Volume layout, routines, approvals and rules. SwiftUI on the Cua Swift SDK.
- [Infinite canvas: every window on one canvas](https://cua.ai/docs/spaces/examples/infinite-canvas.md): A SwiftUI example on the Cua Spaces SDK: a hotkey opens a zoomable canvas of live windows from macOS, Linux and Omarchy Spaces, with agent and coworker cursors and agent threads.
- [Examples in other languages](https://cua.ai/docs/spaces/examples/other-languages.md): Agent-app examples on Cua Spaces in TypeScript, Rust (Tauri) and Swift: named bots, each working on a Space you watch live.

### Reference

- [Reference](https://cua.ai/docs/spaces/reference.md): Every Spaces tool by category, with parameters, results, errors and approval.
- [Space lifecycle](https://cua.ai/docs/spaces/reference/lifecycle.md): Register, provision, list and release Spaces.
- [Commands and files](https://cua.ai/docs/spaces/reference/commands-and-files.md): Run shell commands and move files in and out of a Space.
- [Streams and display](https://cua.ai/docs/spaces/reference/streams-and-display.md): Stream a Space's desktop or windows, and show a Space to the operator.
- [Services](https://cua.ai/docs/spaces/reference/services.md): Discover and call the MCP services inside a Space.
- [Agents](https://cua.ai/docs/spaces/reference/agents.md): Run coding agents inside a Space.
- [Persistent agents](https://cua.ai/docs/spaces/reference/persistent-agents.md): Named agents whose memory outlives their Space: homes in Cua Volume, routines, notifications, pause and resume, access to the user's computers.
- [Session teleport](https://cua.ai/docs/spaces/reference/teleport.md): Move a signed-in app session from this machine into a Space.
- [Keyvault](https://cua.ai/docs/spaces/reference/keyvault.md): Use the user's saved credentials in a Space, only with their approval.
- [Hotspot](https://cua.ai/docs/spaces/reference/hotspot.md): Share this machine's network with a Space.
- [Cua Volume](https://cua.ai/docs/spaces/reference/volume.md): Read and write the volume every Space and agent shares, decide who may reach what, and see where it is stored and how it syncs.
- [Sharing](https://cua.ai/docs/spaces/reference/sharing.md): Let other cua.ai accounts watch or edit a Space, through the relay.
- [Your cloud](https://cua.ai/docs/spaces/reference/cloud.md): Create Spaces in your own AWS, Google Cloud or Modal account: connect it, check it, and clean up what Cua created there.
- [Errors](https://cua.ai/docs/spaces/reference/errors.md): Tool error kinds and JSON-RPC protocol errors of the Spaces tools.

## Cua Fleets

- [Overview](https://cua.ai/docs/fleets.md): Dedicated cloud capacity for sandboxes. Named pools, templates, claims, warm replicas and Terraform.
- [Quickstart](https://cua.ai/docs/fleets/quickstart.md): Your first cloud sandbox and your first pool. Start a desktop in the cloud, verify a file hash, then apply a pool and claim from it.

### Guides

- [Create capacity and claim from it](https://cua.ai/docs/fleets/guides/capacity-and-claims.md): Apply a named pool, claim sandboxes from it, keep a sandbox across processes, and bound lifetimes with TTLs.
- [Keep sandboxes warm](https://cua.ai/docs/fleets/guides/warm-pools.md): Keep ready sandboxes in a managed or dedicated pool so claims bind in seconds, and size warm capacity against cost.
- [Manage pools with Terraform](https://cua.ai/docs/fleets/guides/terraform.md): Declare Linux and Windows Fleet pools as fleets_pool resources, apply them, and export existing pools to Terraform.
- [Publish a Fleet image](https://cua.ai/docs/fleets/guides/images.md): Which runtime boots a cloud image, and how to build, push and reference your own image with cua image build.
- [Add sidecars on Fleet](https://cua.ai/docs/fleets/guides/sidecars.md): Put sidecar containers in a pool's template, and how Fleet runs them next to container and VM sandboxes.
- [Deliver per-claim secrets](https://cua.ai/docs/fleets/guides/claim-secrets.md): Give each claim of a warm pool its own cua-spacesd token, delivered into the sandbox when the claim binds.
- [Troubleshoot cloud sandboxes](https://cua.ai/docs/fleets/guides/troubleshoot.md): Diagnose credential, access, image and readiness errors on Fleet, and clean up sandboxes and pools left behind.

### Reference

- [Reference](https://cua.ai/docs/fleets/reference.md): Cua Fleets objects, REST operations, SDK methods and Terraform schema.
- [Pool](https://cua.ai/docs/fleets/reference/pool.md): Warm capacity: how many sandboxes of one template Fleet keeps ready to claim.
- [Template](https://cua.ai/docs/fleets/reference/template.md): What each sandbox of a pool runs: image, runtime, size, services and probes.
- [Claim](https://cua.ai/docs/fleets/reference/claim.md): A lease on one sandbox from a pool, until it is released or expires.
- [Sandbox](https://cua.ai/docs/fleets/reference/sandbox.md): One running machine of a pool, bound to at most one claim.
- [Secret](https://cua.ai/docs/fleets/reference/secret.md): Secrets the SDK writes in a pool's namespace: a claim's env token and registry pull credentials.
- [Service URL](https://cua.ai/docs/fleets/reference/service-url.md): HTTP and WebSocket access to a sandbox's services, and shareable signed URLs.
- [Image](https://cua.ai/docs/fleets/reference/image.md): A build recipe Fleet turns into a runnable image, and its build status.
- [Namespace](https://cua.ai/docs/fleets/reference/namespace.md): The account-owned scope every pool, template, claim and secret lives in.
- [API key](https://cua.ai/docs/fleets/reference/api-key.md): OAuth client credentials that act for your user, for the SDK, CLI and Terraform.
- [Terraform provider](https://cua.ai/docs/fleets/reference/terraform.md): The trycua/fleets provider: settings, environment variables and resources.
- [Errors](https://cua.ai/docs/fleets/reference/errors.md): Fleet HTTP status codes and SDK errors, with causes and fixes.

## Cua Bench

- [Overview](https://cua.ai/docs/cua-bench.md): Run the Cua Bench registry tasksets against any computer-use agent on local or cloud desktop sandboxes, or write your own verifiable tasks.
- [Quickstart](https://cua.ai/docs/cua-bench/quickstart.md): Run a Cua Bench registry taskset on a local desktop sandbox with one command, read the results, then run it with an agent.

### Guides

- [Run a benchmark](https://cua.ai/docs/cua-bench/guides/benchmarks.md): Pick a taskset from the Cua Bench registry, run it with one cb run command on a local or cloud desktop sandbox, and read the results.
- [Write a task](https://cua.ai/docs/cua-bench/guides/write-a-task.md): The Cua Bench task module: variants, setup, oracle and evaluator, the computer configuration, and the session operations they use.
- [Adapter benchmarks](https://cua.ai/docs/cua-bench/guides/adapter-benchmarks.md): Run established benchmarks through Cua Bench adapters: OSWorld-Verified, MiniWoB++, WebVoyager, Online-Mind2Web, WebGym and the grounding sets, locally or in the cloud.
- [Convert benchmark images](https://cua.ai/docs/cua-bench/guides/convert-benchmark-images.md): Turn an upstream benchmark environment into one Cua image with a container and a VM variant, publish it, and check eval parity.

### Reference

- [Reference](https://cua.ai/docs/cua-bench/reference.md): Every cb command, the task module contract and the result files, generated from cua-bench.

#### CLI

- [CLI](https://cua.ai/docs/cua-bench/reference/cli.md): cua-bench: run computer-use tasks and benchmark datasets on local and cloud sandboxes.
- [cb run](https://cua.ai/docs/cua-bench/reference/cli/run.md): Run a task or dataset with an agent or the oracle, run one interactively, and follow runs.
- [cb task](https://cua.ai/docs/cua-bench/reference/cli/task.md): Inspect, list, scaffold and generate tasks.
- [cb dataset](https://cua.ai/docs/cua-bench/reference/cli/dataset.md): List registry datasets and build training data from runs.
- [cb trace](https://cua.ai/docs/cua-bench/reference/cli/trace.md): View traces and agent trajectories.
- [Sandboxes and account](https://cua.ai/docs/cua-bench/reference/cli/env.md): Cloud pools and local sandboxes, sign-in, status and cleanup.
- [cb agent](https://cua.ai/docs/cua-bench/reference/cli/agent.md): Scaffold custom agents.
- [Images and platforms](https://cua.ai/docs/cua-bench/reference/cli/images.md): The canonical images and platform configurations.
- [Task definition](https://cua.ai/docs/cua-bench/reference/task-definition.md): The task module contract: cb.Task, setup_config, the lifecycle decorators, the session API and actions.
- [Results](https://cua.ai/docs/cua-bench/reference/results.md): The result.json and summary.json a run writes: every key, its type and meaning.

## Lume

- [Overview](https://cua.ai/docs/lume.md): Create and run macOS and Linux VMs on Apple silicon from the CLI, a local HTTP API or MCP.
- [Quickstart](https://cua.ai/docs/lume/quickstart.md): Install Lume, create a vanilla macOS Tahoe VM from an Apple restore image, and connect over SSH.

### Guides

- [Manage VMs](https://cua.ai/docs/lume/guides/manage-vms.md): Run, attach to, share folders with, store, clone, resize and remove Lume VMs, change SIP, and update Lume.
- [Unlock Metal for a workload](https://cua.ai/docs/lume/guides/gpu-passthrough.md): Unlock a tested Metal capability profile for one workload inside a Lume macOS VM.
- [Drive Lume from scripts and agents](https://cua.ai/docs/lume/guides/api-and-mcp.md): Drive Lume from scripts through its local HTTP API, or from an AI client through its MCP server.
- [Run Omarchy and OpenClaw](https://cua.ai/docs/lume/guides/linux-guests.md): Run Omarchy on an ARM64 Arch Linux guest with optional Rosetta, and run an OpenClaw gateway in an isolated Lume VM.

### Concepts

- [How Lume works](https://cua.ai/docs/lume/concepts/how-lume-works.md): Offline unattended setup, SIP through paired Recovery, macOS disk growth with rollback, and what Lume telemetry collects.

### Reference

- [Reference](https://cua.ai/docs/lume/reference.md): Every lume command, HTTP endpoint and MCP tool, generated from the Lume binary.

#### CLI

- [CLI](https://cua.ai/docs/lume/reference/cli.md): A lightweight CLI and local API server to build, run and manage macOS VMs.
- [VMs](https://cua.ai/docs/lume/reference/cli/vms.md): Create, run, inspect, change, clone and delete virtual machines.
- [Images](https://cua.ai/docs/lume/reference/cli/images.md): Pull, push and convert VM images, find restore images, and prune the cache.
- [Guest access](https://cua.ai/docs/lume/reference/cli/guest.md): Prepare unattended setup, open SSH sessions and change SIP in a guest.
- [Server](https://cua.ai/docs/lume/reference/cli/server.md): Run the HTTP API and MCP server, read its logs, and dump the CLI and API definitions.
- [Configuration](https://cua.ai/docs/lume/reference/cli/config.md): Storage locations, the image cache and telemetry settings.
- [Updates](https://cua.ai/docs/lume/reference/cli/updates.md): Check for, apply and choose the channel of Lume updates.
- [HTTP API](https://cua.ai/docs/lume/reference/http-api.md): Every endpoint of the Lume HTTP API server (lume serve).
- [MCP tools](https://cua.ai/docs/lume/reference/mcp-tools.md): Every tool Lume's MCP server exposes, with its parameters.

## Cua CLI

- [Overview](https://cua.ai/docs/cua-cli.md): The cua command: sign-in, sandboxes local and cloud, one-shot computer actions, MCP, Spaces, unattended access and teleport.
- [Quickstart](https://cua.ai/docs/cua-cli/quickstart.md): Install the cua CLI, start a local sandbox, run a command in it, then start the same sandbox in the cloud.

### Guides

- [Authenticate](https://cua.ai/docs/cua-cli/guides/authentication.md): Sign in with cua auth login, see which Fleet credential the CLI uses, create API keys, and authenticate in GitHub Actions.
- [Connect an agent over MCP](https://cua.ai/docs/cua-cli/guides/mcp-server.md): Serve Spaces, sandbox, computer and skills tools to an AI client with cua mcp or cua daemon mcp, and grant only the permissions it needs.

### Reference

- [Reference](https://cua.ai/docs/cua-cli/reference.md): Every cua command and every cua MCP tool, generated from the CLI definition.

#### CLI

- [CLI](https://cua.ai/docs/cua-cli/reference/cli.md): Cua: sandboxes, computer actions, MCP, skills and the cua daemon.
- [cua auth](https://cua.ai/docs/cua-cli/reference/cli/auth.md): Sign in, inspect your Fleet identity, manage API keys and CI tokens.
- [cua sandbox](https://cua.ai/docs/cua-cli/reference/cli/sandbox.md): Create, list, connect to, suspend and delete sandboxes, local or in the cloud.
- [Sandbox access](https://cua.ai/docs/cua-cli/reference/cli/sandbox-access.md): Run commands, copy files, forward ports, take screenshots, reach the MCP and view a sandbox (`cua sandbox exec`, `cp`, `view`, ...).
- [cua config](https://cua.ai/docs/cua-cli/reference/cli/config.md): User defaults: where sandboxes run (`default.on`), the default kind and runtime, cloud defaults.
- [cua viewer](https://cua.ai/docs/cua-cli/reference/cli/viewer.md): One local browser page that lists every sandbox and opens its viewer.
- [cua do](https://cua.ai/docs/cua-cli/reference/cli/do.md): One-shot computer actions (screenshot, click, type, windows) against a selected target.
- [cua image](https://cua.ai/docs/cua-cli/reference/cli/image.md): Pull, build and push OCI images, and manage Fleet image resources; list the image catalog.
- [cua fleet](https://cua.ai/docs/cua-cli/reference/cli/fleet.md): Managed Fleet pools behind `cua sandbox create`.
- [cua mcp](https://cua.ai/docs/cua-cli/reference/cli/mcp.md): The stdio MCP server for AI assistants.
- [cua agent](https://cua.ai/docs/cua-cli/reference/cli/agent.md): Run coding agents (Claude Code, Codex, Gemini CLI, ...) inside a sandbox and follow, continue or stop them.
- [cua agents](https://cua.ai/docs/cua-cli/reference/cli/agents.md): Install cua skills and the cua MCP server into AI coding agents.
- [cua skills](https://cua.ai/docs/cua-cli/reference/cli/skills.md): Record, list and replay demonstrations (skills).
- [cua trajectory](https://cua.ai/docs/cua-cli/reference/cli/trajectory.md): Recorded `cua do` trajectories.
- [cua spaces](https://cua.ai/docs/cua-cli/reference/cli/spaces.md): Registered Spaces and your machines on the relay.
- [cua cloud](https://cua.ai/docs/cua-cli/reference/cli/cloud.md): Your own AWS, Google Cloud or Modal account as a place for sandboxes and Spaces: connect, test, status, sweep.
- [cua host](https://cua.ai/docs/cua-cli/reference/cli/host.md): Set this machine up for unattended access.
- [cua devices](https://cua.ai/docs/cua-cli/reference/cli/devices.md): Enroll this device on the relay, approve, rename and revoke devices, and read the access log.
- [cua teleport](https://cua.ai/docs/cua-cli/reference/cli/teleport.md): Move a desktop app session into a sandbox.
- [cua volume](https://cua.ai/docs/cua-cli/reference/cli/volume.md): Cua Volume: files, versions, grants, access requests and audit of the volume every Space and agent shares.
- [cua keyvault](https://cua.ai/docs/cua-cli/reference/cli/keyvault.md): Set up, unlock and lock the Cua Keyvault, with the OS key store or a passphrase.
- [cua spacesd](https://cua.ai/docs/cua-cli/reference/cli/spacesd.md): Talk to a cua-spacesd directly by URL.
- [cua daemon](https://cua.ai/docs/cua-cli/reference/cli/daemon.md): Start, stop and inspect the cua daemon.
- [cua cache](https://cua.ai/docs/cua-cli/reference/cli/cache.md): See and reclaim the disk used by images, sandboxes, builds and logs.
- [cua runtime](https://cua.ai/docs/cua-cli/reference/cli/runtime.md): Inspect and provision local runtimes.
- [cua doctor](https://cua.ai/docs/cua-cli/reference/cli/doctor.md): Check the host, an image and a sandbox guest, or eval parity between image variants.
- [cua telemetry](https://cua.ai/docs/cua-cli/reference/cli/telemetry.md): Anonymous usage telemetry (status, off, show-last).

#### MCP tools

- [MCP tools](https://cua.ai/docs/cua-cli/reference/mcp-tools.md): Every tool the cua MCP server exposes, with its permission and parameters.
- [Spaces tools](https://cua.ai/docs/cua-cli/reference/mcp-tools/spaces.md): The Spaces contract: register, claim and release Spaces, run commands, move files, stream, run agents and teleport.
- [Image tools](https://cua.ai/docs/cua-cli/reference/mcp-tools/images.md): The sandbox image catalog: which images exist and which ship a browser.
- [Sandbox tools](https://cua.ai/docs/cua-cli/reference/mcp-tools/sandbox.md): Create and manage sandboxes, local or in the cloud.
- [Teleport tools](https://cua.ai/docs/cua-cli/reference/mcp-tools/teleport.md): Move a signed-in browser session, per site, into a sandbox through the Cua Keyvault, with consent.
- [Computer tools](https://cua.ai/docs/cua-cli/reference/mcp-tools/computer.md): Screenshots, input, windows, files and shell on a sandbox desktop.
- [Skills tools](https://cua.ai/docs/cua-cli/reference/mcp-tools/skills.md): List, read, record and delete skills (recorded demonstrations).

## More documentation

- [Docs for agents](https://cua.ai/docs/docs-for-agents.md): Give a coding agent Cua's docs and source as Markdown or over MCP.