Focusrelaymcp, developed by Deverman, is an OmniFocus MCP server built with Swift, JavaScript, and Shell. It operates locally on Mac systems, enabling AI assistants to interact with OmniFocus data through documented Omni Automation APIs. This program allows users to safely read, update, complete, and organize tasks and projects directly within OmniFocus. Distributed as a native Swift binary via Homebrew, Focusrelaymcp provides essential tools for managing task workflows and ensuring data integrity. It supports various AI clients, making it a key component for automated task management.

What does Focusrelaymcp connect to?

Focusrelaymcp connects directly to OmniFocus, a task management application, functioning as an OmniFocus MCP server. It operates entirely on the user's local Mac, establishing communication with a bridge plug-in within OmniFocus. This plug-in gains access to OmniFocus data using its documented Omni Automation APIs. Importantly, Focusrelaymcp does not transmit OmniFocus data to any FocusRelay-operated cloud service, maintaining local data residency. Its primary purpose is to mediate interactions between AI assistants and OmniFocus tasks.

Core features and tools

Focusrelaymcp exposes specific callable tools designed for interaction with OmniFocus data. These include preview and verification, which are described as MCP tool arguments. The preview tool is designed to resolve IDs and validate proposed changes without saving them to OmniFocus. The verification tool runs after OmniFocus saves data, reading back affected values and reporting any mismatches as failures. Focusrelaymcp supports standard stdio transport protocols and is compatible with AI clients such as Codex, Claude Code, and OpenCode.

Who is it for?

Focusrelaymcp is for users who want to empower AI assistants to safely interact with OmniFocus tasks and projects using documented Omni Automation APIs. This includes reading, updating, completing, and organizing tasks. Its local operation ensures data privacy, making it suitable for those concerned about cloud data exposure. The guided setup process helps users integrate the system efficiently. Skip this if you require a hosted cloud service for your task management AI interactions.

How to get started

The authentication model for Focusrelaymcp involves local, on-device operation, so no credentials are required to connect to a hosted cloud service. It supports standard stdio transport protocols for communication with AI clients.

Use cases

• Processing an inbox in manageable batches with a guided workflow that starts with unresolved captures, asks for confirmation before changes, and verifies approved updates.
• Keeping ordinary inbox cleanup focused on unfinished captures without mixing in completed or dropped history.
• Important changes where users can ask the AI assistant to "preview first, then apply with verification".

Focusrelaymcp provides a secure, on-device solution for integrating AI assistants with OmniFocus data. Its local operation ensures privacy, and specialized tools like preview and verification support controlled, validated task management. The program is compatible with several AI clients and offers guided workflows for common task organization needs. This enables users to automate and refine their OmniFocus usage effectively. Try processing an inbox in manageable batches with a guided workflow that starts with unresolved captures.