M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
A

Agent Device

Community
Let AI agents inspect, control, and verify apps.
How to install ↓ Compare GitHub source repository ↗
Category
Dev Tools #133 of 438
Stars
★ 5.1k Very popular
Transport
stdio (local process)
Runtime
Node.js 22.12+
Credentials
No credential needed
License
MIT
Last commit
59FMRS · C

An open-source MIT-licensed device automation CLI from Callstack for agent-driven app inspection, control, verification, debugging, and evidence capture. The manifest lists version 0.20.0, an npm package, and stdio transport; the source does not provide specific MCP tool identifiers or a client configuration example.

Strongest · Documentation 15/20 Weakest · Security and permissions 8/20

Reliability
10/20
Security and permissions
8/20
Maintenance
14/20
Documentation
15/20
Setup experience
12/20
Why each score
Reliability 10/20
The manifest declares an npm package, stdio transport, and version, while the README describes sessions, an inspect-act-verify workflow, and platform dependencies, making the normal path plausible. However, the supplied material contains no reviewable MCP initialization/tool-list implementation, committed tests, or error-handling evidence. Handshake success and consistency between declared and actual tools therefore remain unverified, so the static calibration requires a deduction.
Security and permissions 8/20
The server can inspect, control, and debug real devices and applications, and may save screenshots, videos, logs, network data, and performance information; its side effects and potentially sensitive data boundary are broad. The material does not explain permissions, credential handling, network data flows, dangerous-operation confirmation, or retention. No explicit credential theft, covert exfiltration, or irreversible-operation red line is evidenced, but the missing security controls require substantial deductions.
Maintenance 14/20
The repository is not archived and provides an MIT license, identifiable GitHub ownership, npm version 0.20.0, a CI badge, and a contribution entry point, indicating an identifiable maintainer and release path. The supplied material does not establish commit cadence, issue response, dependency updates, a security-response channel, or release history, so a near-full score is not justified.
Documentation 15/20
The README supplies prerequisites, basic commands, examples, supported platforms, capability descriptions, operating model, CI/cloud use cases, limitations, and documentation links, giving it good breadth. Evidence is still missing for authentication, complete MCP client configuration, detailed tool parameters and error semantics, cost, permission boundaries, and troubleshooting. The linked external documentation was not reviewed, so points are deducted.
Setup experience 12/20
Global npm installation, Node.js requirements, the doctor check, and help entry points provide a clear starting path, with local, CI, and cloud deployment modes described. There is no concrete configuration example for a mainstream MCP client, and device/platform preparation across iOS, Android, TV, and desktop is complex. Important setup details also depend on unreviewed external documentation, so setup is only moderate.

Static review · not runListed 2026-07-29

Read the FMRS scoring method →

Fit and risk

What it can accessRuns commands or codeControls a browser

Best for

  • Teams building React Native, Expo, Flutter, or native apps
  • Teams validating apps on local devices, simulators, emulators, or CI
  • QA and development workflows where an agent chooses actions at runtime

Not for

  • Teams seeking a complete replacement for conventional maintained test suites
  • Workflows requiring physical Fire TV, capture, or Vega selector backends
  • Users without an available supported device, simulator, emulator, or desktop target

Required permissions

  • Access to and control of connected iOS, Android, TV, desktop, or other supported device environments
  • Reading accessibility trees, logs, network data, performance information, and crash details where supported
  • Saving screenshots, videos, traces, and other verification evidence when requested

Risks and side effects

  • Device actions can change application state or perform test operations
  • Element targeting and verification may be less reliable when accessibility data is poor
  • Earlier @eN references may no longer identify the same element; use references from the latest output
  • Vega OS support is currently VVD-only; physical Fire TV, capture, and selector backends are unsupported

Setup

Before you start

Runtime：Node.js 22.12+

Node.js 22.12 or newer is required; web automation requires Node.js 24 or newer. Install the CLI with npm install -g agent-device@latest, then run agent-device doctor to check setup. The server manifest declares npm package agent-device version 0.20.0 with stdio transport.

Check that it works

Run `npm install -g agent-device@latest`, then run `agent-device doctor`; a passing setup report confirms the CLI is installed and targets are configured.

Troubleshooting

  1. Run agent-device doctor first to check setup
  2. Verify that the Node.js version meets the requirements for the target
  3. Run agent-device help workflow for current CLI behavior and workflow guides
  4. Check that the target device, simulator, or emulator is running and accessible
  5. Prefer references from the latest snapshot or settle diff; use screenshots or video when accessibility data is insufficient

Things to try

Once connected, you can ask your AI assistant things like:

  • Open the Contacts app on the iOS simulator and take a screenshot of the current form
  • Tap the login button in my Android app and verify it navigated to the home screen
  • Use agent-device to take an accessibility snapshot of the React Native screen I just changed
  • Save the steps from this session as an .ad script so I can replay it in CI

Use cases

Let coding agents operate running apps and verify code changes
Run mobile end-to-end tests and repeatable workflows
Diagnose issues with screenshots, logs, traces, and performance data
Save .ad scripts or export strict Maestro YAML for replay

Overview

agent-device is a command-line tool that lets AI agents inspect, control, debug, and verify apps on iOS, Android, tvOS, Android TV, Amazon Vega OS TV, the web, macOS, and Linux. It supports accessibility snapshots, element targeting, device actions, screenshots, video, logs, traces, network data, performance samples, crash details, and React profiles where supported.

Similar servers

Context7 80 · B

Upstash's official server providing up-to-date third-party library docs for AI coding assistants

★ 63.3k · Tools 2 Compare with this →
Compare with other MCP servers

Source revision 3d9190107e75 Data synced 2026-10-19 Read the FMRS scoring method