Things3 MCP Server

Things3 MCP Server – Overview

The Things3 MCP server bridges the gap between macOS‑only task management and AI assistants that follow the Model Context Protocol. By exposing a rich set of 25 tools, it allows an AI client to perform almost every action available in the native Things3 app—creating, updating, deleting, and organizing tasks, projects, areas, and tags—all through a single, standardized interface. This eliminates the need for custom scripts or manual workarounds, enabling developers to build intelligent productivity workflows that can read from and write to a user’s Things3 database without exposing the underlying AppleScript or URL schemes.

At its core, the server solves a common pain point for developers who want to automate or augment task management: Things3’s macOS‑only, AppleScript‑centric API is difficult to integrate into cross‑platform AI agents. The MCP server abstracts that complexity, providing a clean JSON‑based contract that any compliant client can consume. Developers benefit from consistent error handling, automatic tag creation, and bulk operations that reduce round‑trip latency. The server also includes intelligent error correction—such as auto‑filling missing titles or resolving date conflicts—ensuring that the assistant’s actions result in valid Things3 items.

Key capabilities include:

• Full CRUD for TODOs, Projects, Areas, and Tags – Each entity type has dedicated tools that support creation, listing, updating, completion toggling, and deletion.
• Hierarchical Tag Management – Tags can be created on‑the‑fly, deleted, or applied in bulk to multiple items, mirroring Things3’s native tagging system.
• Bulk Operations – Move or update large sets of items with a single call, improving performance and reducing API calls.
• Logbook Search – Retrieve completed items within a date range, useful for reporting or analytics.
• Performance Optimizations – Connection pooling and AppleScript caching keep latency low, even when handling many items.

Typical use cases include:

• AI‑powered task assistants that can read a user’s inbox, suggest priorities, or automatically move items to projects based on context.
• Contextual knowledge bases where an AI can fetch recent completed tasks to inform future recommendations.
• Cross‑platform integrations, such as syncing Things3 with other tools (e.g., calendar, email) via an MCP‑compliant bridge.
• Automated workflows that trigger on events (e.g., new email) and create corresponding Things3 tasks, complete them when a deadline passes, or generate summary reports.

Integration is straightforward for MCP clients: configure the server’s command and authentication token, then call the exposed tools with JSON payloads. Because the server adheres to MCP’s standard of declaring resources, prompts, and sampling, developers can combine it with other MCP services—such as language models or database connectors—to build sophisticated, end‑to‑end productivity solutions.