Skip to content
MCP server

SAIR

Run Android instrumented tests on real phones or lease a device from an agent

Runs locallyStatus and history
Install
Pricing
Freemium
Platforms
CLI, API
Runs
Locally
Version
0.1.36
Listing updated
3 days ago

About SAIR

SAIR is a freemium MCP server for developer tools. It runs locally on your machine and is started by your AI client.

SAIR started as a way for CI jobs to share physical Android phones without colliding: a job takes a lock, waiting jobs queue, and ADB traffic is routed through a proxy. The MCP server, sair-mcp, extends that to coding agents so they can test on real hardware without having adb or a proxy on their own machine.

In one call an agent can upload an app APK and a test APK and get back per-test results, instrument output and a logcat tail. Or it can lease a phone, install an APK, run shell commands, take screenshots and read logcat. Each lease length depends on plan: 15 minutes on Free, up to 4 hours on Enterprise.

It talks to SAIR's agent API using a key from the dashboard, ships as a Go binary and as an MCP bundle for Claude Desktop, and the same service has a GitHub Action for ordinary CI. SAIR positions itself against Firebase Test Lab and AWS Device Farm on cost, and the pricing page compares them.

Install

Run in a terminal

claude mcp add sair --env SAIR_API_KEY='<SAIR_API_KEY>' -- npx -y https://github.com/compscidr/sair/releases/download/v0.1.36/sair-mcp.mcpb

Add to claude_desktop_config.json

{
  "mcpServers": {
    "sair": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/compscidr/sair/releases/download/v0.1.36/sair-mcp.mcpb"
      ],
      "env": {
        "SAIR_API_KEY": "<SAIR_API_KEY>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "sair": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/compscidr/sair/releases/download/v0.1.36/sair-mcp.mcpb"
      ],
      "env": {
        "SAIR_API_KEY": "<SAIR_API_KEY>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "sair": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/compscidr/sair/releases/download/v0.1.36/sair-mcp.mcpb"
      ],
      "env": {
        "SAIR_API_KEY": "<SAIR_API_KEY>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "sair": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/compscidr/sair/releases/download/v0.1.36/sair-mcp.mcpb"
      ],
      "env": {
        "SAIR_API_KEY": "<SAIR_API_KEY>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "sair": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/compscidr/sair/releases/download/v0.1.36/sair-mcp.mcpb"
      ],
      "env": {
        "SAIR_API_KEY": "<SAIR_API_KEY>"
      }
    }
  }
}

Configuration

  • SAIR_API_KEYAn API key from https://sair.run/dashboardRequired
  • SAIR_API_URLAgent API base URL (default: https://api.sair.run)

Tools this server provides (7)

Taken from the project documentation; the running server may differ.

run_tests
Uploads app and test APKs and returns per-test results, instrument output and a logcat tail.
list_devices
Lists phones the key may use.
lease_device
Takes a phone and returns a lease id.
install_apk
Installs an APK on the leased phone.
shell
Runs a command and returns output and exit code.
screenshot
Returns a PNG of the screen.
logcat
Returns recent logcat lines.

What it does

One-call test runs
Upload app and test APKs and receive per-test results and logcat.
Device leasing
Lease a phone for installs, shell commands, screenshots and logs.
No local adb
The agent machine needs neither a proxy nor adb.
Device locking
Queues and locks stop CI jobs from interfering on the same phone.
GitHub Action
Drop-in acquire and release steps around gradlew connectedCheck.
Failure visibility
Jobs link to their CI run with ADB commands and optional screenshots or video.

Good for

  • Let a coding agent verify a change on a real phone
  • Run connectedCheck in CI on phones you own
  • Reproduce a bug by installing an APK and reading logcat
  • Share a few test phones across many repositories
  • Replace per-minute device cloud billing

Pricing

Freemium

Free for 2 devices and 10 jobs/day; Solo, Team and Enterprise tiers (prices not published for Solo)

Prices change. Check sair.run before you buy.

SAIR FAQ

What does it need to run?

A SAIR_API_KEY from the dashboard; SAIR_API_URL is optional.

How is it installed?

With go install, a release tarball, or the sair-mcp.mcpb bundle for clients like Claude Desktop.

Is there a free tier?

Yes: 2 devices, 10 jobs a day and agent leases up to 15 minutes.

Is a self-hosted option offered?

The Enterprise plan lists a self-hosted option.

Is SAIR free?

SAIR has a free plan, with paid plans for more usage or features.

How do I add SAIR to Claude Code?

Run this command in your terminal: claude mcp add sair --env SAIR_API_KEY='<SAIR_API_KEY>' -- npx -y https://github.com/compscidr/sair/releases/download/v0.1.36/sair-mcp.mcpb

What are the best alternatives to SAIR?

The closest alternatives to SAIR are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to SAIR

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools