io.github.oozoofrog/baepsae MCP Server
io.github.oozoofrog/baepsae
Local MCP server for iOS Simulator and macOS app automation with native Swift bridge
What is the io.github.oozoofrog/baepsae MCP server?
The baepsae MCP server is a local automation tool for iOS Simulator and macOS applications, providing a TypeScript MCP layer backed by a native Swift bridge. It enables UI inspection, input automation, app lifecycle management, and system interaction through unified generic tools.
baepsae lets you automate iOS Simulator and macOS apps through accessibility APIs and native frameworks. Use it to inspect UI hierarchies, tap buttons, type text, launch/terminate apps, take screenshots, record videos, and control system features—all from Claude, Cursor, or other MCP clients. It's designed for testing, automation workflows, and interactive app control.
How to install io.github.oozoofrog/baepsae
Copy-paste configuration for popular MCP clients.
Tools & capabilities
Tools this server exposes to the agent.
analyze_ui— Inspect the accessibility tree and UI structure of the target appquery_ui— Search the UI hierarchy by text, identifier, or other criteriatap— Tap a UI element by accessibility identifier or labeltap_tab— Switch tabs by index with semantic-first fallback to geometrytype_text— Type text into the focused element via keyboard or pasteswipe— Perform swipe gestures on the targetscroll— Scroll within a UI containerdrag_drop— Drag and drop UI elementswait_for_ui— Wait for a UI element to appeardetect_dialog— Detect and identify dialog boxesread_ui_value— Read the current value of a UI elementset_ui_value— Set the value of a UI elementread_ui_param— Read parameters or attributes of a UI elementhit_test— Perform hit testing at a specific coordinateenumerate_ui— List all UI elements matching criteriakey— Press a single keykey_sequence— Execute a sequence of key presseskey_combo— Press multiple keys simultaneouslytouch— Perform touch input at coordinatesinput_source— Manage input source selection
Use cases
- Automate iOS app testing by launching apps, tapping UI elements, and verifying screen content
- Control macOS applications programmatically to test workflows or perform repetitive tasks
- Inspect and interact with app UIs through accessibility APIs without writing native code
- Record and stream simulator video for debugging or documentation
- Manage simulator lifecycle: list devices, install/uninstall apps, and capture screenshots
io.github.oozoofrog/baepsae MCP server FAQ
baepsae is an MCP server that automates iOS Simulator and macOS apps using accessibility APIs and native Swift bridges. It provides 43 tools for UI inspection, input automation, app management, and system control.
Yes, baepsae is open-source and free to use. Install via npm: `npm install -g mcp-baepsae` or run directly with `npx mcp-baepsae@latest`.
baepsae requires macOS 14+, Xcode/Xcode Command Line Tools, and Node.js 18+. It does not support Linux or Windows because it depends on macOS-specific frameworks (AppKit, CoreGraphics, Accessibility).
Run `bash scripts/install.sh --tool all` to register baepsae with all supported clients, or use `claude mcp add baepsae -- npx -y mcp-baepsae@latest` for manual setup. The installer handles configuration for Claude Code, Claude Desktop, Codex, Gemini, and other clients.
Yes. Accessibility permission must be granted to the automation host/runtime process (e.g., `node`, your terminal, or MCP client app), not the target app. Use System Settings > Privacy & Security > Accessibility to enable the process, then restart it before retrying.
No, baepsae only supports iOS Simulator and macOS apps. It uses `xcrun simctl` and Accessibility APIs that are specific to the simulator and macOS environments.
README (reference)
Source of truth, from the repository.
mcp-baepsae
<p align="center"> <img src="assets/baepsae.png" width="300" alt="baepsae"> </p>Baepsae (Vinous-throated Parrotbill) — A tiny Korean bird. Round, chubby, and constantly hopping around chirping. Known for its grit — even when a little bird tries to keep up with a stork, it never gives up. This project is small too, but it pecks away at your simulators tirelessly.
Local MCP server for iOS Simulator and macOS app automation with a TypeScript MCP layer and a Swift native bridge.
한국어 문서는 README-KR.md를 참고하세요.
Table of Contents
- Prerequisites
- Platform Support
- Install
- Permissions
- MCP Setup (Recommended)
- Client Matrix
- For LLM
- Manual Setup (Fallback)
- Project Structure
- Commands
- MCP Tool Status
- Usage Examples
- Troubleshooting
Prerequisites
- macOS 14+
- Xcode + iOS Simulator
- Node.js 18+
- Swift 6+
Platform Support
| Platform | Supported | Notes |
|---|---|---|
| macOS | Yes | Primary platform. Required for iOS Simulator and Accessibility APIs. |
| Linux | No | Native binary depends on AppKit, CoreGraphics, and Accessibility frameworks. |
| Windows | No | Native binary depends on AppKit, CoreGraphics, and Accessibility frameworks. |
Why macOS only?
The Swift native bridge (baepsae-native) uses macOS-specific frameworks (AppKit, CoreGraphics, Accessibility) to interact with iOS Simulator and macOS applications. These frameworks are not available on Linux or Windows. The TypeScript MCP layer also relies on xcrun simctl, which is part of Xcode Command Line Tools and only available on macOS.
Requirements summary:
- macOS 14 or later -- required for iOS Simulator automation and Accessibility API access.
- Xcode or Xcode Command Line Tools -- required for Swift 6+ compilation of the native binary and
xcrun simctlcommands. - Node.js >= 18.0.0 -- required to run the TypeScript MCP server.
Permissions
Accessibility permission is required for UI inspection and input automation features (use unified generic tools such as analyze_ui, tap, right_click).
The important detail is that permission usually needs to be granted to the automation host / runtime process, not to the target app you are automating.
Which process usually needs permission?
- Direct native binary invocation
- Example:
baepsae-native ... - Most relevant entry: the
baepsae-nativebinary itself, plus the terminal/shell app that launched it
- Example:
- Node / npx MCP runtime
- Example:
node dist/index.js,npx -y mcp-baepsae@latest - Most relevant entry: the runtime process (
node), plus the terminal or MCP client app that launched it
- Example:
- Desktop / CLI MCP clients
- Example: Claude Code, Codex CLI/Desktop, Gemini CLI
- Relevant entries can include the MCP client app, the terminal host, and the runtime process depending on launch path
Recommended setup flow
- Open System Settings > Privacy & Security > Accessibility.
- Enable the terminal or MCP client app you actually use.
- Enable the runtime process if listed (
node,bun, etc.). - For direct native invocation, also enable the
baepsae-nativebinary entry if it appears separately. - If an entry is missing, click
+and add it manually.
Important note
After granting permission, the launching process may need to be restarted before macOS applies the change.
If the error persists, quit and relaunch the terminal, MCP client, or runtime process that started mcp-baepsae.
For simulator targets, selector-based actions (tap / right_click with id or label) search in-app content by default. Set all: true to include Simulator chrome UI.
Install
Option A) npm (easiest)
# Run directly without installing
npx mcp-baepsae@latest
# Or install globally
npm install -g mcp-baepsae
On macOS, the Swift native binary is built automatically during install. If Swift is not available, the server still works with simctl-based features.
If you plan to use UI inspection or input automation right away, grant Accessibility permission to the terminal / MCP client / runtime process you will actually launch the server with.
Option B) From source
git clone https://github.com/oozoofrog/mcp-baepsae.git
cd mcp-baepsae
npm install
npm run build
MCP Setup (Recommended)
Use the installer script that calls each client's MCP install flow directly.
bash scripts/install.sh --tool all
Client Matrix
| Client | Install Path | Script Target | Notes |
|---|---|---|---|
| Claude Code | claude mcp add | --tool claude-code | Supports multi-scope via --claude-scope |
| Claude Desktop | claude mcp add --scope user | --tool claude-desktop | User-level registration |
| Codex CLI | codex mcp add | --tool codex-cli | Also works for Codex Desktop settings |
| Codex Desktop | Codex CLI MCP config | --tool codex-desktop | Uses same MCP registry as CLI |
| OpenCode | Global config (~/.config/opencode/opencode.json) | --tool opencode | Auto-updated by installer |
| Gemini | gemini mcp add | --tool gemini | User scope setup |
| Google Antigravity | Gemini-compatible MCP flow | --tool antigravity | Uses Gemini MCP command path |
| GitHub Copilot | copilot or gh copilot session | --tool copilot | Interactive/session-based setup |
For LLM
If you are an LLM agent setting up this MCP server, here is everything you need:
Before trying UI automation tools, remember that Accessibility permission usually belongs to the host/runtime process (
node, terminal, MCP client), not the target app under automation.
Quick start (one command)
# Register baepsae as an MCP server for all supported clients
bash scripts/install.sh --tool all
If you installed via npm instead of cloning the repo, use npx:
# Claude Code
claude mcp add baepsae -- npx -y mcp-baepsae@latest
# Codex CLI
codex mcp add baepsae -- npx -y mcp-baepsae@latest
Automation flags
# Preview commands without executing
bash scripts/install.sh --tool all --dry-run
# Verify environment and dependencies
bash scripts/install.sh --tool all --doctor
# Unregister from all clients
bash scripts/install.sh --tool all --uninstall
Runtime options
The installer supports multiple runtimes via --runtime:
| Flag | Command | When to use |
|---|---|---|
--runtime node (default) | node dist/index.js | Local source build |
--runtime npx | npx -y mcp-baepsae@latest | npm registry, no global install |
--runtime bunx | bunx mcp-baepsae@latest | Bun users |
--runtime global | mcp-baepsae | After npm install -g mcp-baepsae |
Manual Setup (Fallback)
Use this when you do not want to run scripts/install.sh.
Using npx (recommended for npm users)
# Claude Code
claude mcp add baepsae -- npx -y mcp-baepsae@latest
# Codex CLI
codex mcp add baepsae -- npx -y mcp-baepsae@latest
# Gemini CLI
gemini mcp add --scope user --transport stdio baepsae npx -y mcp-baepsae@latest
When using npx, the relevant Accessibility entry is commonly the spawned node runtime plus the terminal / MCP client that launched it.
Using local build
# Claude Code (project)
claude mcp add --scope project --env="BAEPSAE_NATIVE_PATH=/ABS/PATH/native/.build/release/baepsae-native" baepsae -- node /ABS/PATH/dist/index.js
# Codex CLI
codex mcp add baepsae --env BAEPSAE_NATIVE_PATH=/ABS/PATH/native/.build/release/baepsae-native -- node /ABS/PATH/dist/index.js
# Gemini CLI
gemini mcp add --scope user --transport stdio -e BAEPSAE_NATIVE_PATH=/ABS/PATH/native/.build/release/baepsae-native baepsae node /ABS/PATH/dist/index.js
When using a local build, check permission on both the runtime (node) and the app that launched it.
If you invoke baepsae-native directly for debugging, check permission on the native binary entry itself as well.
Project Structure
- MCP server entry point:
src/index.ts - Tool modules:
src/tools/(info, simulator, ui, input, media, system) - Shared utilities:
src/utils.ts,src/types.ts - Native binary entry point:
native/Sources/main.swift - Native command handlers:
native/Sources/Commands/ - Native binary output:
native/.build/release/baepsae-native - TS tests:
tests/mcp.contract.test.mjs,tests/unit.test.mjs,tests/mcp.real.test.mjs - Swift tests:
native/Tests/BaepsaeNativeTests/
Commands
npm run build # Build TypeScript + native Swift binary
npm test # Contract/integration tests
npm run test:real # Real simulator smoke test (requires booted simulator)
npm run test:real:preflight # Environment diagnostics only
npm run test:real:sim # iOS simulator phases only (skips Phase 4)
npm run test:real:mac # macOS Safari phase only
npm run verify # test + test:real
npm run setup:mcp # Alias for scripts/install.sh
MCP Tool Status
43 tools implemented end-to-end.
Official public MCP surface: unified generic tools
The public API surface is intentionally single-scheme: use unified generic tools with a target argument, rather than sim_* / mac_* names.
| Category | Tools |
|---|---|
| UI | analyze_ui, query_ui, tap, tap_tab, type_text, swipe, scroll, drag_drop, wait_for_ui, detect_dialog, read_ui_value, set_ui_value, read_ui_param, hit_test, enumerate_ui |
| Input | key, key_sequence, key_combo, touch, input_source, list_input_sources |
| Workflow | run_steps |
| System | list_windows, activate_app, screenshot_app, right_click, focus_window, context_menu_action, watch_notification |
| Simulator-only | list_simulators, screenshot, record_video, stream_video, open_url, install_app, launch_app, terminate_app, uninstall_app, button, gesture |
| macOS/system | list_apps, menu_action, get_focused_app, clipboard |
| Utility | baepsae_help, baepsae_version, doctor |
Target routing is explicit in the arguments: udid for simulator, bundleId / appName for macOS.
type_text policy
type_text accepts exactly one input source: text, stdinText, or file.
method: "auto"resolves to:pastefor simulator targetskeyboardfor macOS targets
method: "paste"uses the simulator pasteboard for simulator targets and a temporary host clipboard replace/restore flow for macOS targets.method: "keyboard"always types character-by-character.
When paste is used, simulator targets update the simulator pasteboard without touching the host clipboard, while macOS targets temporarily overwrite the host clipboard and restore it after submission. Successful responses report the input source, target kind, requested method, used method, paste transport, and any auto fallback that was applied.
tap_tab policy
tap_tab is semantic-first, geometry-last.
- First it tries actionable descendants exposed under the tab bar.
- If SwiftUI/Simulator does not expose real tab button descendants, it can use a semantic proxy row (for example, app-provided top navigation buttons that map cleanly to the same tabs).
- Only if neither semantic path exists does it fall back to tab-bar geometry.
This matters because SwiftUI TabView in Simulator may expose the tab bar as a generic AXGroup text="Tab Bar" without actionable child buttons.
Usage Examples
Switch tabs by index with semantic-first fallback:
// Prefer this when your app exposes a real tab bar but individual tab buttons
// are not consistently addressable via query_ui/tap selectors.
tap_tab({ udid: "...", index: 1, tabCount: 3 })
Unified simulator app accessibility quickstart (inside app UI):
// 1) Launch your app in the target simulator
launch_app({ udid: "...", bundleId: "com.example.app" })
// 2) Inspect or search accessibility tree (in-app content scope by default)
analyze_ui({ udid: "..." })
query_ui({ udid: "...", query: "Login" })
// 3) Interact by accessibility identifier/label
tap({ udid: "...", id: "login-button" })
// Optional: include Simulator chrome/system UI in selector lookup
tap({ udid: "...", label: "Home", all: true })
Open a URL (iOS Simulator):
// Open Naver Mobile
open_url({ udid: "...", url: "https://m.naver.com" })
Manage Apps (iOS Simulator):
// Install an app
install_app({ udid: "...", path: "/path/to/App.app" })
// Launch Safari
launch_app({ udid: "...", bundleId: "com.apple.mobilesafari" })
// Terminate Safari
terminate_app({ udid: "...", bundleId: "com.apple.mobilesafari" })
macOS App Automation:
// List running macOS apps
list_apps({})
// Take screenshot of a macOS app
screenshot_app({ bundleId: "com.apple.Safari" })
Troubleshooting
Accessibility permission checklist
-
The permission target is usually the automation host/runtime process, not the target app.
-
Run
doctorfirst to inspect host process, parent process, native binary, booted simulator availability, and accessibility readiness in one place. -
Check the error message for:
- current host process
- parent process
- inferred launch mode
-
If you launched through
npx/node, grant permission to the runtime and the launching terminal / MCP client. -
If you launched
baepsae-nativedirectly, grant permission to the native binary entry and the launching terminal / shell app. -
After changing permission, restart the launching process before retrying.
-
Invalid environment variable formaton Claude setup:- Use current script (
scripts/install.sh) orclaude mcp add --env="KEY=value" ...format.
- Use current script (
-
Missing native binaryerror:- Run
npm run buildand confirmnative/.build/release/baepsae-nativeexists.
- Run
-
Accessibility permission error is ambiguous:
- Current versions include host/parent process diagnostics and inferred launch mode in the error text so you can see which executable path likely needs permission.
-
Real smoke test diagnostics:
- Run
npm run test:real:preflightto print environment and capability diagnostics without executing the full suite. - Run
npm run test:real:simto focus on simulator-capability coverage, ornpm run test:real:macfor the macOS Safari subset.
- Run
-
Multiple Simulator windows are open and selectors hit the wrong device:
- Current versions scope simulator selectors to the target
udidwindow first. - If you intentionally need Simulator chrome/system UI, use
all: true.
- Current versions scope simulator selectors to the target
-
OpenCode does not show
baepsae:- Re-run
bash scripts/install.sh --tool opencode --skip-install --skip-buildand check~/.config/opencode/opencode.json.
- Re-run
-
Copilot not auto-registered:
- Copilot MCP flow is interactive/session-based. Re-run installer with
--interactive.
- Copilot MCP flow is interactive/session-based. Re-run installer with
-
Real smoke test skipped:
- Boot an iOS simulator first, then run
npm run test:real.
- Boot an iOS simulator first, then run
Related MCP servers
Search 324k icons or generate new ones in the Lucide grammar - grammar-locked lint+critique+repair

Apuchat
Real-time chat for AI agents. Claude Code, Cursor, Cline and Codex join channels over MCP.
Discover, evaluate, and compare SaaS APIs. Agent-verified scores and metrics.
View repository →
RogerRat
Real-time chat hub for AI agents — Claude Code, Cursor, Cline, Codex over MCP or REST.

RogerThat
Real-time chat hub for AI agents — Claude Code, Cursor, Cline, Codex over MCP or REST.
Preços, atas, contratos e sanções das APIs públicas de compras do governo brasileiro
View repository →