Apple Doc MCP: SwiftUI & UIKit Documentation for Cursor & Claude

Apple Doc MCP is a Model Context Protocol server that integrates Apple’s official developer documentation directly into your AI coding environment. Whether you are querying SwiftUI views, searching for UIKit controllers, or exploring the Foundation class hierarchy, this tool retrieves real-time data from Apple and formats it into clean Markdown. This ensures your Large Language Model (LLM) receives high-quality, relevant context without unnecessary noise.

Advanced Search. Full wildcard support allows you to use * and ? to locate specific symbols across every Apple framework.

Framework Exploration. Navigating the structures of SwiftUI, UIKit, Foundation, and other core libraries is straightforward and intuitive.

Comprehensive Documentation. Access full symbol descriptions, method signatures, and relevant code examples.

Granular Filtering. Narrow your results by platform (iOS, macOS, watchOS, tvOS), symbol type, or specific framework.

Real-Time Data. The server remains synchronized with Apple’s latest published documentation.

Update Alerts. The server automatically notifies you when the repository contains new changes or data improvements.

AI-Ready Output. Information is delivered in sanitized Markdown, allowing LLMs to parse and implement the documentation efficiently.

Installation

1. Clone the repository:

   git clone github.com/MightyDillah/apple-doc-mcp.git cd apple-doc-mcp

2. Setup is minimal. The pre-built distribution is located in the dist/ directory, meaning no manual build step is required. You simply need to point your MCP client to the included files.

Quick Setup

Step 1: Configure Your AI Assistant

Depending on your preferred environment, edit the following configuration file:

• Claude Desktop: ~/.config/claude/claude_desktop_config.json
• Cursor: Settings (Cmd/Ctrl + ,) → Extensions → MCP
• Continue.dev: ~/.continue/config.json
• VS Code (Claude Dev): Settings → MCP Servers

Add the following configuration block:

{
  "mcpServers": {
    "apple-doc-mcp": {
      "command": "node",
      "args": ["/path/to/apple-doc-mcp/dist/index.js"]
    }
  }
}

Note: Ensure you use the absolute file path to the directory where you cloned the repository.

Step 2: Restart and Verify

Restart your AI assistant. To verify the connection, ask: "List available Apple technologies." You should see four new tools active in the environment. Upon launch, the server also checks for updates and will notify you if a new version is available.

How to Use

Interact with your assistant as you normally would. It will invoke the Apple Doc MCP tools automatically when it needs technical context.

Browse Available Technologies

"Use apple-doc-mcp to list all current Apple frameworks."
"Retrieve the latest available Apple technologies from the documentation."
"Search Apple documentation for all supported frameworks."

Explore a Framework

"Browse the SwiftUI framework structure using apple-doc-mcp."
"Fetch current UIKit topics from the Apple docs."
"Search Apple documentation for details regarding the Foundation framework."

Search for Specific APIs

"Search Apple's SwiftUI docs for drag-and-drop APIs."
"Use apple-doc-mcp to find RPBroadcast* classes in ReplayKit."
"Find all *View* symbols within Apple frameworks."
"Locate all *Controller* classes in UIKit using the Apple docs."

Access Detailed Documentation

"Fetch the latest SwiftUI View protocol documentation from Apple."
"Use apple-doc-mcp to find the UIViewController documentation."
"Search Apple's current docs for NSURLSession details."

Available Tools

list_technologies
Provides a comprehensive list of every Apple framework and technology.

get_documentation
Retrieves in-depth documentation for a specific symbol or framework. The tool automatically detects whether the path refers to a technology or a specific symbol.

Parameters:

• path (required): The documentation path, such as "documentation/SwiftUI/View", or a framework name like "SwiftUI".

search_symbols
Performs a filtered search across all Apple frameworks.

Parameters:

• query (required): The search term (supports * and ? wildcards).
• framework (optional): Limits the search to a single framework.
• symbolType (optional): Filters by type, such as class, protocol, or struct.
• platform (optional): Limits results to specific platforms like iOS or macOS.
• maxResults (optional): Sets the maximum number of results (defaults to 20).

check_updates
Scans the git repository for available updates. This runs automatically at startup and requires no manual parameters.

Troubleshooting

The server fails to start:

• Ensure you are running Node.js version 18 or higher.
• Double-check the path in your MCP configuration; it must point specifically to dist/index.js.
• Verify that your JSON syntax is correct (watch for missing commas or quotes).
• Restart your AI assistant or the host application after making any configuration changes.

No tools are displayed in the assistant:

• This usually indicates the server failed to launch.
• Confirm that you are using an absolute path rather than a relative one.
• Test the server manually in your terminal by running: node /path/to/apple-doc-mcp/dist/index.js.
• Note that no build step is required, as the dist folder is provided in the repository.

No search results found:

• Try using a broader query. For example, use "*View*" instead of a literal "View".
• Remove specific filters (like platform or symbol type) to increase the search scope.

Initial search is slow:

• The server generates a local cache during the first search to improve performance. Subsequent queries will be significantly faster.
• If speed is a concern, try reducing the maxResults parameter.