sim-use

sim-use is a cross-platform CLI tool that gives AI agents the ability to observe and act on iOS Simulator and Android screens.

Sim-use is a cross-platform command-line interface that drives Apple accessibility APIs, the iOS Simulator HID pipeline, and Android AccessibilityService through a single command surface. It turns any device screen into a token-efficient outline that large language models can read and reason about in a few hundred tokens. Agents can tap elements using aliases, element IDs, labels, or coordinates without needing raw JSON trees. A background daemon amortizes initialization costs to complete each observe-act round trip in approximately three hundred milliseconds. The tool supports iOS simulators, Android emulators, physical Android devices, and experimental physical iOS devices.

sim-use at a glance

Best for: Software engineer
Runs on: macOS, iOS, Android, CLI
Pricing: Free
Released: June 2026
Category: AI coding tool
Website: github.com/lycorp-jp/sim-use

Key features

Observe screen state into compact outlines
Tap elements by alias, ID, label, or coordinate
Run batch command chaining for iOS
Stream video and record screen actions
Input text using type or IME-safe paste
Drive hardware buttons and key sequences
Probe software keyboard state

Use it for

Verify mobile app builds automatically with an AI agent
Execute automated test scripts across iOS and Android
Inspect accessibility trees and UI hierarchies
Perform multi-step mobile interactions in agent loops

Alternatives to sim-use

Simulator Use — It runs tests and reads screens fast with a token-efficient native engine.
agent-device — Token-efficient snapshots and semantic references for native mobile and TV apps.
serve-sim — Test and interact with hosted Apple Simulators directly from a web browser.
iphone-use — Operates real iPhones via HTTP and MCP without touching the host screen.
Mobile MCP — Connects AI agents directly to real mobile devices for automated testing and tasks.
Phone Harness — Connects Claude Code or Codex to real iOS and Android devices.
AgentDock — Manages multiple coding agents from one screen-edge inbox without switching tabs.
Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
AG-UI — Standardizes agent-user interactions and generative UI for applications.
jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
OpenCode — Works with 75+ model providers, local models, and existing subscriptions
Codeg — Run multiple AI coding agents together and manage them from mobile devices.

All sim-use alternatives