Applescript MCP Server

The Applescript‑MCP server bridges the gap between large language models and macOS by exposing a rich, AppleScript‑based API through the Model Context Protocol. Developers can now let AI assistants perform native system tasks—such as creating calendar events, manipulating the clipboard, or sending notifications—without leaving the conversational flow. This eliminates the need for custom shell scripts or third‑party automation tools, making it straightforward to build end‑to‑end workflows that blend AI reasoning with platform capabilities.

At its core, the server offers a collection of high‑level commands grouped into intuitive categories. The Calendar feature lets an assistant add or list events, while the Finder commands provide file selection and search functionality. System controls cover volume adjustments, dark mode toggling, and retrieving the front‑most application. Messaging is handled through Messages (send, list, search) and Mail (compose, read), enabling the assistant to manage communication directly. The Clipboard module gives read/write access, allowing the AI to copy data into or out of the system clipboard seamlessly. Even terminal interactions are supported via iTerm, giving AI access to command‑line workflows when needed.

Real‑world scenarios that benefit from this server include automated meeting scheduling, where an assistant can pull calendar availability, create events, and notify participants. Developers building productivity tools can let the AI read selected files from Finder, summarize their contents, and copy key excerpts to the clipboard. A mobile‑friendly workflow might involve sending a system notification or toggling Do Not Disturb based on user context. The server’s design also supports iterative refinement: an AI can query the state of a task (e.g., “What’s on my inbox?”), act on it, and then confirm the result.

Integration with existing MCP‑based pipelines is seamless. An AI client sends a structured request to the server, receives a JSON response with the result or error, and can chain multiple calls in a single conversational turn. Because the server is written in Node.js and targets macOS 10.15+, it runs natively on any modern Mac, requiring only the standard Node runtime and user permission for AppleScript execution. Its modular command set allows developers to extend or restrict capabilities per application, ensuring security and privacy while still offering powerful automation.