# MobiLoop

MobiLoop gives an AI coding agent guarded tools for mobile builds, Appium runs, security checks, and evidence-based failure reports.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 11
- Author: enessubass
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/enessubass/mobiloop-mcp
- Updated: 2026-10-15
- HTML: https://affpapa.org/en/nevibecode/mcp/mobiloop

## What it does

MobiLoop connects an MCP client to a real mobile development setup. It can inspect and patch a workspace, build the app, install it on an Android or iOS target, drive it with Appium, and check logs, screenshots, page source, and API responses. The result is a report backed by tool output rather than an agent claim.

The server separates Android and iOS tools, stores screen checkpoints, replays stable setup paths, and runs JSON flows with waits, taps, typing, assertions, and evidence steps. It also scans Flutter, React Native, Android, and iOS source for possible end-to-end scenarios. High-impact tools use an approval gate, while common secrets are redacted from text artifacts.

MobiLoop does not replace the Android SDK, Xcode, Appium, or platform drivers. iOS simulator work requires macOS, and Appium must be running before Appium or flow-replay tools are used.

## Requirements
- Needs: Docker
- Environment variables: `MOBILOOP_WORKSPACE_ROOT`, `APPIUM_SERVER_URL` — keys and tokens: ask the human for the values, never commit them or put them in shared files.

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add --env MOBILOOP_WORKSPACE_ROOT=<MOBILOOP_WORKSPACE_ROOT> --env APPIUM_SERVER_URL=<APPIUM_SERVER_URL> --transport stdio mobiloop -- docker run -i --rm -e MOBILOOP_WORKSPACE_ROOT -e APPIUM_SERVER_URL ghcr.io/enessubass/mobiloop-mcp:0.1.0-alpha.15
```
Replace <…> with your own values

### Codex

**Platform CLI**
```bash
codex mcp add mobiloop --env MOBILOOP_WORKSPACE_ROOT=<MOBILOOP_WORKSPACE_ROOT> --env APPIUM_SERVER_URL=<APPIUM_SERVER_URL> -- docker run -i --rm -e MOBILOOP_WORKSPACE_ROOT -e APPIUM_SERVER_URL ghcr.io/enessubass/mobiloop-mcp:0.1.0-alpha.15
```
Replace <…> with your own values

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "mobiloop": {
            "command": "docker",
            "args": [
                "run",
                "-i",
                "--rm",
                "-e",
                "MOBILOOP_WORKSPACE_ROOT",
                "-e",
                "APPIUM_SERVER_URL",
                "ghcr.io/enessubass/mobiloop-mcp:0.1.0-alpha.15"
            ],
            "env": {
                "MOBILOOP_WORKSPACE_ROOT": "<MOBILOOP_WORKSPACE_ROOT>",
                "APPIUM_SERVER_URL": "<APPIUM_SERVER_URL>"
            }
        }
    }
}
```
Replace <…> with your own values

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "mobiloop": {
            "type": "stdio",
            "command": "docker",
            "args": [
                "run",
                "-i",
                "--rm",
                "-e",
                "MOBILOOP_WORKSPACE_ROOT",
                "-e",
                "APPIUM_SERVER_URL",
                "ghcr.io/enessubass/mobiloop-mcp:0.1.0-alpha.15"
            ],
            "env": {
                "MOBILOOP_WORKSPACE_ROOT": "<MOBILOOP_WORKSPACE_ROOT>",
                "APPIUM_SERVER_URL": "<APPIUM_SERVER_URL>"
            }
        }
    }
}
```
Replace <…> with your own values

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add -e MOBILOOP_WORKSPACE_ROOT=<MOBILOOP_WORKSPACE_ROOT> -e APPIUM_SERVER_URL=<APPIUM_SERVER_URL> mobiloop docker run -i --rm -e MOBILOOP_WORKSPACE_ROOT -e APPIUM_SERVER_URL ghcr.io/enessubass/mobiloop-mcp:0.1.0-alpha.15
```
Replace <…> with your own values

## When to use it
- build and install a mobile app on a target
- run an Appium UI flow with captured evidence
- collect logs, screenshots, and page source
- classify failures from runtime evidence
- scan mobile settings for security issues

## Example prompts
- Build the mobile app, install it on a target, and run an Appium check with evidence.
- Scan the Flutter and Android source for candidate end-to-end scenarios.
- Compare logs, screenshots, and API responses, then classify the failure and write a report.

## Similar
- [appium-skill](https://affpapa.org/en/nevibecode/skills/appium-skill.md) — skill · 422 installs
- [Ghost in the Droid](https://affpapa.org/en/nevibecode/mcp/ghost-in-the-droid.md) — MCP server · ★ 400
- [capacitor-testing](https://affpapa.org/en/nevibecode/skills/capacitor-testing.md) — skill · 1300 installs
- [firebase-apk-scanner](https://affpapa.org/en/nevibecode/skills/firebase-apk-scanner.md) — skill · 5400 installs
- [momentic-mobile-test](https://affpapa.org/en/nevibecode/skills/momentic-mobile-test.md) — skill · 54300 installs
- [argent-device-interact](https://affpapa.org/en/nevibecode/skills/argent-device-interact.md) — skill · 26300 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md