MobiLoop MCP

Guarded MCP servers for agentic mobile build-test-fix loops.

Documentation site | Security model | Tool reference

code change -> build -> install on device -> Appium test -> evidence -> classify -> report

MobiLoop MCP is a controlled tool layer between an AI coding agent and a real mobile development environment. It lets an MCP client read and patch a mobile project, build it, install it on Android or iOS targets, drive the app through Appium, verify logs/screens/API results, remember known app-flow checkpoints, and produce evidence-based reports.

The name reflects the core contract: mobile work should run through a measurable loop of change, build, device execution, verification, and evidence-backed triage.

It is built for the workflow where the agent does not just write code. It builds, runs, tests, observes, classifies failures, and hands back evidence. An agent can still use the separate guarded code tools to patch and retest, but that patch step is intentionally outside the default orchestrator.

Today, MobiLoop provides guarded build-test-verify loops and evidence-based failure classification. Fully automated patch-and-retest is intentionally kept outside the default orchestrator until stricter approval, rollback, and review controls are enabled.

Highlights

• Evidence-first mobile loops: build logs, screenshots, Appium XML source, logcat/simulator logs, API responses, screenshot diffs, and iteration records.
• Android and iOS tool split: Android adb/emulator tools and iOS xcrun simctl/xcodebuild tools are separated.
• Appium UI automation: semantic taps, typing, swipes, back navigation, visibility assertions, screenshots, and accessibility summaries.
• Flow memory: record runtime screen checkpoints, remember the latest passing path, and auto-replay stable setup steps to a target checkpoint.
• Scenario generation and flow DSL: scan source for candidate E2E scenarios, then run high-level JSON flows with wait/tap/type/assert/evidence steps.
• Source-flow analysis: scan Flutter, React Native, Android, and iOS source for screen, route, transition, and visible-text candidates.
• Root-cause classification: classify logcat evidence into app bugs, automation errors, missing environment, remote rules, and test-data issues.
• Server-side approval gate: secure mode requires valid approval payloads for high-impact tools.
• Built-in mobile security loop: scan source and platform settings, generate a test plan, compare fixes, and gate release decisions without another scanner package.
• Redaction by default: redact common secrets, bearer tokens, API keys, emails, and phone numbers from text artifacts, command output, and MCP/CLI text responses.
• Guarded code tools: workspace-only reads/searches/patches, forbidden secret paths, guarded branches, commits, and PR creation.
• Docker-ready MCP runtime: package the Node MCP server in Docker while keeping mobile SDKs, emulators, devices, and Appium on the host or runner.
• Composable binaries: run everything as one server or split each responsibility into its own MCP server.

What This Is

This project provides MCP tools for this architecture:

AI / MCP client
  |
  v
MobiLoop MCP
  |
  |-- code tools
  |-- environment preflight
  |-- build tools
  |-- Android device tools
  |-- iOS simulator tools
  |-- Appium tools
  |-- verification tools
  |-- flow-memory replay tools
  |-- loop/report tools
  |-- CI publication tools
  |-- Android/iOS orchestrators
  |-- security scan and release gate tools
  |
  v
mobile repo + emulator/device + Appium + build toolchain

The server does not claim that a test passed because a model says so. A pass should be backed by tool output: command exit codes, screenshots, page source, log checks, API assertions, and recorded loop iterations.

What This Is Not

• It is not a replacement for Android SDK, Xcode, Flutter, Gradle, React Native, Appium, or platform drivers.
• It is not a universal mobile emulator container. iOS simulator requires macOS, and Android emulator portability depends on host acceleration and device access.
• It is not an unrestricted shell bridge. Tools are structured and guarded.
• It is not a production deployer. Release signing, store upload, and production secrets remain outside the default scope.

Requirements

Install only what your target app needs.

Workflow	Host	Required tools
MCP runtime	macOS, Linux, Windows	Node.js 20+
Android build/test	macOS, Linux, Windows	Android SDK, adb, emulator or physical device, Java/Gradle as needed, Appium 2, UiAutomator2 driver
Flutter Android	macOS, Linux, Windows	Flutter SDK, Android SDK, Appium for UI flows
React Native Android	macOS, Linux, Windows	Node/npm, Android Gradle toolchain, Android SDK, Appium
iOS simulator	macOS only	Xcode, xcrun simctl, iOS Simulator, Appium 2, XCUITest driver
Docker MCP runtime	macOS, Linux, Windows	Docker, plus host-side mobile tools when driving devices

Start Appium before Appium or flow replay tools:

appium --address 127.0.0.1 --port 4723

Local Appium installs also work:

npx appium --address 127.0.0.1 --port 4723

For Android, make sure the Appium process can see:

export ANDROID_HOME=/absolute/path/to/android/sdk
export ANDROID_SDK_ROOT=/absolute/path/to/android/sdk
export PATH="$ANDROID_HOME/platform-tools:$ANDROID_HOME/emulator:$PATH"

For an already-running Genymotion device, add its Android SDK platform tools to PATH, then verify the device before starting a flow:

export PATH="$HOME/Library/Android/sdk/platform-tools:$PATH"
adb devices -l

Install From Source

npm ci
npm test

Run the all-in-one MCP server:

MOBILOOP_WORKSPACE_ROOT=/absolute/path/to/mobile/app \
node dist/src/index.js

For development:

npm run dev

CLI Fallback

When an MCP client cannot expose the server as callable tools, use the CLI wrapper:

MOBILOOP_WORKSPACE_ROOT=/absolute/path/to/mobile/app \
node dist/src/cli.js list-tools

Inspect tool policy metadata:

MOBILOOP_WORKSPACE_ROOT=/absolute/path/to/mobile/app \
node dist/src/cli.js list-tools --json

The same metadata is available inside MCP through policy.list_tools.

Call any tool directly:

MOBILOOP_WORKSPACE_ROOT=/absolute/path/to/mobile/app \
node dist/src/cli.js call flow.generate_test_scenarios '{"goal":"login smoke and validation"}'

Generate scenario candidates:

MOBILOOP_WORKSPACE_ROOT=/absolute/path/to/mobile/app \
node dist/src/cli.js generate-scenarios "cover onboarding, login, and validation"

MCP Client Configuration

All-In-One Server

Use this for local development and simpler MCP clients.

{
  "mcpServers": {
    "mobiloop": {
      "command": "node",
      "args": ["/absolute/path/to/mobiloop-mcp/dist/src/index.js"],
      "env": {
        "MOBILOOP_WORKSPACE_ROOT": "/absolute/path/to/mobile/app",
        "APPIUM_SERVER_URL": "http://127.0.0.1:4723"
      }
    }
  }
}

Split Servers

Use split servers when you want tighter policy boundaries per responsibility.

{
  "mcpServers": {
    "mobile-code": {
      "command": "node",
      "args": ["/absolute/path/to/mobiloop-mcp/dist/src/servers/code.js"],
      "env": {
        "MOBILOOP_WORKSPACE_ROOT": "/absolute/path/to/mobile/app"
      }
    },
    "mobile-build": {
      "command": "node",
      "args": ["/absolute/path/to/mobiloop-mcp/dist/src/servers/build.js"],
      "env": {
        "MOBILOOP_WORKSPACE_ROOT": "/absolute/path/to/mobile/app"
      }
    },
    "mobile-device": {
      "command": "node",
      "args": ["/absolute/path/to/mobiloop-mcp/dist/src/servers/device.js"],
      "env": {
        "MOBILOOP_WORKSPACE_ROOT": "/absolute/path/to/mobile/app"
      }
    },
    "mobile-appium": {
      "command": "node",
      "args": ["/absolute/path/to/mobiloop-mcp/dist/src/servers/appium.js"],
      "env": {
        "MOBILOOP_WORKSPACE_ROOT": "/absolute/path/to/mobile/app",
        "APPIUM_SERVER_URL": "http://127.0.0.1:4723"
      }
    },
    "mobile-flow": {
      "command": "node",
      "args": ["/absolute/path/to/mobiloop-mcp/dist/src/servers/flow.js"],
      "env": {
        "MOBILOOP_WORKSPACE_ROOT": "/absolute/path/to/mobile/app",
        "APPIUM_SERVER_URL": "http://127.0.0.1:4723"
      }
    },
    "mobile-verify": {
      "command": "node",
      "args": ["/absolute/path/to/mobiloop-mcp/dist/src/servers/verify.js"],
      "env": {
        "MOBILOOP_WORKSPACE_ROOT": "/absolute/path/to/mobile/app"
      }
    },
    "mobile-loop": {
      "command": "node",
      "args": ["/absolute/path/to/mobiloop-mcp/dist/src/servers/loop.js"],
      "env": {
        "MOBILOOP_WORKSPACE_ROOT": "/absolute/path/to/mobile/app"
      }
    }
  }
}

All binaries are listed below.

Binary	Scope
mobiloop	CLI wrapper for listing tools, calling tools, and generating scenarios
mobiloop-mcp	All tools
mobiloop-code-mcp	Code and git tools
mobiloop-env-mcp	Environment preflight and compatibility matrix
mobiloop-build-mcp	Dependency, lint, test, and APK build tools
mobiloop-device-mcp	Android adb and emulator tools
mobiloop-ios-mcp	iOS simulator and Xcode tools
mobiloop-appium-mcp	Appium UI automation tools
mobiloop-verify-mcp	Assertions and evidence collection
mobiloop-flow-mcp	Source-flow analysis and checkpoint replay
mobiloop-loop-mcp	Iteration records and reports
mobiloop-ci-mcp	Artifact manifests, GitHub summaries, PR comments
mobiloop-orchestrator-mcp	Android and iOS build-install-test-verify loops
mobiloop-security-mcp	Static mobile security scan, test plan, scan comparison, release gate

Configuration

Secure mode does not read configuration from the project directory. Set workspace and Appium values in the host environment:

export MOBILOOP_WORKSPACE_ROOT=/absolute/path/to/mobile/app
export APPIUM_SERVER_URL=http://127.0.0.1:4723

For a host-controlled file configuration, copy the example and point to it explicitly:

cp mobiloop.config.example.json mobiloop.config.json

Or point to it explicitly:

export MOBILOOP_CONFIG=/absolute/path/to/mobiloop.config.json

AGENTIC_MOBILE_MCP_CONFIG and AGENTIC_MOBILE_WORKSPACE_ROOT are still accepted as legacy fallbacks, but new projects should use the MOBILOOP_* names.

Common fields:

Field	Default	Purpose
securityMode	secure	Secure ignores project-local config; trusted enables explicit overrides
workspaceRoot	current working directory	Mobile app workspace the MCP server may access
artifactsDir	.mobiloop	Evidence, logs, screenshots, reports, flow memory
runId	unset	Optional run identifier; writes artifacts under .mobiloop/runs/<id>
maxCommandMs	120000	Default command timeout
maxOutputBytes	1048576	Output cap for command tools
maxFixAttempts	3	Suggested fix-loop limit
maxTestIterations	5	Orchestrator loop limit
maxRuntimeMinutes	30	Suggested total runtime limit
allowedBranchPattern	^feature/ai-[A-Za-z0-9._/-]+$	Branches where commit tools are allowed
appiumServerUrl	http://127.0.0.1:4723	Trusted-mode Appium endpoint; secure mode uses host environment
adbPath	adb	Android Debug Bridge path
emulatorPath	emulator	Android emulator CLI path
xcrunPath	xcrun	iOS simulator CLI path
xcodebuildPath	xcodebuild	Xcode build CLI path
sqlitePath	sqlite3	SQLite CLI path for read-only assertions
apiAllowlist	localhost only	URLs allowed for API verification
appiumAllowlist	localhost only	Trusted-mode Appium origin allowlist
forbiddenPathGlobs	secret-like defaults	Files blocked from read/write operations
toolPolicies	built-in defaults	Trusted-mode per-tool risk and approval metadata overrides
requireApproval	true in secure mode	Require approval payloads for high-impact tools
redactArtifacts	true	Redact common secrets and PII from text artifacts and text responses

MOBILOOP_ARTIFACTS_DIR may point to a dedicated host-controlled evidence mount, such as /artifacts in the read-only Docker security server.

Environment variables override selected fields:

export MOBILOOP_WORKSPACE_ROOT=/absolute/path/to/mobile/app
export APPIUM_SERVER_URL=http://127.0.0.1:4723
export MOBILOOP_RUN_ID=local-login-smoke
export MOBILOOP_REQUIRE_APPROVAL=true
export MOBILOOP_SECURITY_MODE=secure

The config schema is available at schema/mobiloop.config.schema.json. See docs/CONFIGURATION.md.

Approval payloads use this shape:

{
  "approval": {
    "approved": true,
    "approvedBy": "human-or-ci",
    "reason": "Run Android validation on emulator",
    "expiresAt": "2026-05-11T12:00:00Z"
  }
}

Recommended First Run

1. Point MOBILOOP_WORKSPACE_ROOT at your mobile app.
2. Run security.scan_source and security.generate_test_plan.
3. Start an emulator or connect a device, then start Appium.
4. Run env.preflight and flow.analyze_from_code.
5. Run approved build/lint/unit-test actions.
6. Install the app, create an Appium session, and verify one small user flow.
7. Collect evidence; if a bug is confirmed, patch, rerun, and compare the security scan.
8. Apply security.release_gate before calling a fix ready for release.

Live Appium Proof

v0.1.0-alpha.10 was validated against an installed MiniTakip Android application on a Genymotion Galaxy S24 running Android 15. The non-mutating proof used MobiLoop to discover the ADB target, create an Appium 3 UiAutomator2 session, capture a screenshot and page-source XML, read the accessibility tree, then close the session. It did not enter form data, create records, or delete application data.

Direct W3C capabilities are accepted by appium.create_session as shown below. MobiLoop wraps them in the WebDriver capabilities envelope before sending them to Appium:

{
  "capabilities": {
    "alwaysMatch": {
      "platformName": "Android",
      "appium:automationName": "UiAutomator2",
      "appium:udid": "127.0.0.1:6555",
      "appium:appPackage": "com.example.app",
      "appium:appActivity": ".MainActivity",
      "appium:noReset": true
    }
  },
  "approval": {
    "approved": true,
    "approvedBy": "human-or-ci",
    "reason": "Open an Appium session for a bounded validation run"
  }
}

After creating a session, use appium.observe_screen and appium.get_accessibility_tree for evidence, then always call appium.delete_session. A successful session only proves the selected screen and assertions; it does not claim that an entire product journey has passed.

For a Flutter Android app, the rough tool sequence is:

env.preflight { "target": "flutter" }
security.scan_source
security.generate_test_plan
flow.analyze_from_code
build.detect_project
build.install_dependencies
build.run_lint
build.run_unit_tests
build.build_debug_apk
device.list_devices
device.install_app
appium.create_session
appium.wait_for_visible
appium.tap_by_text
verify.assert_no_crash_in_logcat
verify.collect_evidence
loop.record_iteration
loop.generate_report

Android Orchestrator

orchestrator.run_android_validation_loop runs a bounded Android loop across build, install, Appium, verification, evidence, and iteration records.

Minimal shape:

{
  "goal": "Build and verify login flow on Android.",
  "kind": "flutter",
  "packageName": "com.example.app",
  "serial": "emulator-5554",
  "runLint": true,
  "runUnitTests": true,
  "buildDebugApk": true,
  "clearAppData": true,
  "collectEvidence": true,
  "maxTestIterations": 3,
  "appiumCapabilities": {
    "platformName": "Android",
    "appium:automationName": "UiAutomator2",
    "appium:deviceName": "Android Emulator",
    "appium:udid": "emulator-5554",
    "appium:appPackage": "com.example.app",
    "appium:appActivity": ".MainActivity",
    "appium:noReset": false
  },
  "appiumSteps": [
    {
      "tool": "appium.wait_for_visible",
      "args": {
        "locator": { "strategy": "text", "value": "Login" },
        "timeoutMs": 10000
      }
    }
  ],
  "expectedTexts": ["Home"]
}

See examples/android-validation-loop.json.

iOS Orchestrator

orchestrator.run_ios_validation_loop runs a bounded iOS simulator loop across xcodebuild, simulator boot, app install/launch, Appium XCUITest, verification, evidence, and iteration records.

Minimal Flutter shape:

{
  "goal": "Build and verify login flow on iOS.",
  "kind": "flutter",
  "workspace": "ios/Runner.xcworkspace",
  "scheme": "Runner",
  "configuration": "Debug",
  "sdk": "iphonesimulator",
  "destination": "platform=iOS Simulator,name=iPhone 15",
  "buildSettings": {
    "ARCHS": "arm64",
    "EXCLUDED_ARCHS": ""
  },
  "simulatorDevice": "iPhone 15",
  "bundleId": "com.example.app",
  "runLint": true,
  "runUnitTests": true,
  "buildIosApp": true,
  "bootSimulator": true,
  "installApp": true,
  "launchApp": true,
  "collectEvidence": true,
  "maxTestIterations": 2,
  "appiumCapabilities": {
    "platformName": "iOS",
    "appium:automationName": "XCUITest",
    "appium:deviceName": "iPhone 15",
    "appium:bundleId": "com.example.app",
    "appium:noReset": false
  },
  "appiumSteps": [
    {
      "tool": "appium.wait_for_visible",
      "args": {
        "locator": { "strategy": "text", "value": "Login" },
        "timeoutMs": 15000
      }
    }
  ],
  "expectedTexts": ["Home"]
}

See examples/flutter-ios-validation-loop.json and docs/QUICKSTART_FLUTTER.md.

buildSettings and xcodebuildArgs are forwarded to ios.build_app, so projects can handle host-specific simulator requirements such as Apple Silicon arm64 simulator builds or custom DerivedData settings without leaving the MCP loop.

AI-Generated Scenario Candidates

MobiLoop can generate candidate E2E scenarios from the app source so the agent starts from a concrete test plan instead of an empty screen.

{
  "tool": "flow.generate_test_scenarios",
  "args": {
    "goal": "cover onboarding, login, form validation, and main navigation",
    "maxScenarios": 8,
    "includeNegativeCases": true
  }
}

The output includes priorities, candidate steps, assertions, source references, and limitations. Treat these as executable candidates: the agent should run them through Appium, collect evidence, and refine them into stable checkpoint paths.

For scripted execution without writing custom client code:

{
  "tool": "flow.run_script",
  "args": {
    "sessionId": "APPIUM_SESSION_ID",
    "steps": [
      { "action": "observe", "waitForAnyText": ["Giriş Yap", "Login"], "waitForPackageIdle": true },
      { "action": "tapText", "text": "Giriş Yap", "matchMode": "auto" },
      {
        "action": "type",
        "locator": { "strategy": "text", "value": "E-posta" },
        "text": "test@example.com",
        "mode": "sendKeys"
      },
      { "action": "assertText", "text": "Ana Sayfa" },
      { "action": "collectEvidence", "label": "login-result" }
    ]
  }
}

Flow Memory And Auto-Replay

Flow memory makes repeated mobile tests faster without pretending that setup screens passed.

current Appium source
  -> normalized screen signature
  -> match recorded checkpoint
  -> replay known semantic actions
  -> arrive at target checkpoint
  -> continue test-specific assertions

The screen signature is built from Appium page source:

• visible text
• accessibility labels and content descriptions
• resource ids
• class names
• clickable text

This data is persisted under:

.mobiloop/flow/memory.json

With runId enabled, the same file lives under .mobiloop/runs/<runId>/flow/memory.json.

Record Checkpoints

At a stable screen:

{
  "testName": "onboarding-to-login",
  "name": "Onboarding 1",
  "order": 1,
  "sessionId": "APPIUM_SESSION_ID",
  "actionToNext": {
    "tool": "appium.tap_by_text",
    "args": {
      "text": "Next"
    }
  }
}

At the next screen:

{
  "testName": "onboarding-to-login",
  "name": "Login",
  "order": 2,
  "sessionId": "APPIUM_SESSION_ID"
}

Record the passing path:

{
  "testName": "onboarding-to-login",
  "status": "passed",
  "checkpointIds": ["onboarding-to-login-001-onboarding-1", "onboarding-to-login-002-login"]
}

Replay Later

Plan without executing:

{
  "sessionId": "APPIUM_SESSION_ID",
  "testName": "onboarding-to-login",
  "targetCheckpointId": "onboarding-to-login-002-login",
  "dryRun": true
}

Execute: