# mobile-next/mobile-mcp

Model Context Protocol Server for Mobile Automation and Scraping (iOS, Android, Emulators, Simulators and Real Devices)

Repository: https://github.com/mobile-next/mobile-mcp
Canonical: https://ross.abutalabs.com/products/mobile-mcp
Homepage: https://mobilenext.ai
Language: TypeScript
License: Apache-2.0
License Family: permissive
Topics: android, ios, mcp, mobile, agent, emulator, physical, real, simulator
Last push: 2026-08-26T14:57:30+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 96, longevity 37
- inputs: {"age_days": 523, "days_push": 7, "days_rel": 24, "gap_med": 5, "n_releases_24m": 52}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6033, forks 515 (observed 2026-08-28T04:09:34.397242+00:00)

## What it is
An open-source Model Context Protocol (MCP) server that lets AI agents and LLMs control iOS and Android devices—emulators, simulators, and real hardware—through accessibility snapshots and screenshot-based interactions. It is part of the Mobile Next ecosystem alongside a CLI, a Playwright-style framework, and a paid cloud device farm.

## Use cases
- automate native iOS and Android app testing with an AI agent
- let Claude or Cursor tap and swipe on a real phone via MCP
- extract data from mobile apps without manual device control
- run multi-step user journeys on simulators and emulators driven by an LLM
- script form filling and data entry on mobile devices
- give coding agents device access to verify app changes with screenshots

## When to choose
- you want an MCP-compatible agent to drive local iOS/Android devices without platform-specific glue
- you need accessibility-tree-based automation that avoids vision-model token costs
- you want one interface across emulators, simulators, and real devices
- you are building agent workflows that include mobile app interaction or scraping

## When to avoid
- you need a traditional scripted test framework without an LLM agent (consider Appium or the companion Mobilewright)
- you only target web browsers rather than native mobile apps
- you require guaranteed SLA support from the open-source project alone
- your environment cannot run Node.js or connect to physical/virtual devices

## Facets
- artifact type: service
- maturity: active
- function: mcp, agent-framework, testing, workflow-automation, web-scraping
- domain: mobile-development, testing, developer-tools
- platform: cli, cross-platform
- tags: mcp-server, mobile-automation, device-farm, llm-agents, appium-alternative, accessibility-tree, simulator, emulator, real-devices, ai-agents, automation, android, ios, nodejs

## Member repositories
- mobile-next/mobile-mcp (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:34.397242+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:49:08.656849+00:00, confidence not recorded.
  - readme: https://github.com/mobile-next/mobile-mcp (fetched 2026-08-28T04:09:34.397242+00:00, sha 9e06eec03419)
  - homepage: https://mobilenext.ai (fetched 2026-08-29T08:45:38.416614+00:00, sha e6c20fe03432)
  - site_page: https://mobilenext.ai/docs (fetched 2026-08-29T08:45:38.420951+00:00, sha 9418cef2cf29)
  - site_page: https://mobilenext.ai/pricing (fetched 2026-08-29T08:45:38.419266+00:00, sha 85acd4655f92)
- Data as of 2026-08-30T08:39:29.467469+00:00.