adb-mcp

An MCP server (TypeScript/Node) that bridges an AI MCP client to Android devices via ADB, exposing tools for device management, app installation/management, logcat, file transfer, UI screenshot/dump, and shell command execution.

Evaluated Mar 30, 2026 (21d ago)
Homepage ↗ Repo ↗ DevTools mcp adb android automation devtools uiautomator claude cursor-ai testing
⚙ Agent Friendliness
53
/ 100
Can an agent use this?
🔒 Security
42
/ 100
Is it safe for agents?
⚡ Reliability
30
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
78
Documentation
70
Error Messages
0
Auth Simplicity
55
Rate Limits
10

🔒 Security

TLS Enforcement
40
Auth Strength
30
Scope Granularity
20
Dep. Hygiene
60
Secret Handling
70

Security posture is primarily determined by local execution and ADB authorization. There is no documented TLS/auth model for the MCP server itself and no documented fine-grained scoping/allowlisting for tools; since it can execute shell commands and manage apps/files, agent-side restrictions are important. Dependency hygiene appears plausible but cannot be verified from provided data. Secret handling guidance is not documented.

⚡ Reliability

Uptime/SLA
0
Version Stability
55
Breaking Changes
30
Error Recovery
35
AF Security Reliability

Best When

You control the host environment and the connected devices (e.g., local dev/emulator/your own lab devices) and want agent-driven ADB operations through MCP.

Avoid When

You cannot restrict what an agent can execute (e.g., arbitrary shell/shell-like actions) or you require strong security controls beyond local device access.

Use Cases

  • Automated QA workflows with an AI agent interacting with Android devices
  • Installing APKs to emulators/devices and verifying app behavior
  • Capturing screenshots/UI hierarchy for UI understanding
  • Collecting logcat output to debug app issues
  • Pushing/pulling files to/from a device under agent control

Not For

  • Running untrusted/unsafe commands on production devices without safeguards
  • Environments requiring network-restricted or air-gapped operation without device access planning
  • Use cases needing strong enterprise auth/audit or multi-tenant separation at a service-provider level

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: No API/service authentication described; relies on local execution and ADB device authorization (USB/network debugging authorization).
OAuth: No Scopes: No

No service-level auth mechanism is documented; access appears to be controlled by who can run the local MCP server and by the underlying ADB authorization state (e.g., 'unauthorized' requires accepting debugging authorization on-device).

Pricing

Free tier: No
Requires CC: No

No hosting/pricing described; appears to be a self-hosted open-source package (MIT).

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • ADB operations can be stateful and device/environment dependent (USB authorization, network connectivity, emulator lifecycle).
  • Tool set includes potentially high-impact actions (e.g., arbitrary shell command execution and app/package management); an agent needs allowlisting/guardrails to avoid destructive behavior.
  • Server must be kept running in a terminal session per README; clients must manage process lifecycle correctly.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for adb-mcp.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered