FocusRelayMCP by Deverman connects AI assistants to OmniFocus on macOS for conversational task access and automation. The tool translates natural language into MCP queries, letting AI clients retrieve schedules, project statuses, and inbox items with time-aware filters. Key functions include project health detection, timezone handling, and JXA-powered communication for fast responses. It targets OmniFocus power users and GTD practitioners who want to inspect and prioritize tasks through an AI client instead of the app UI.

What tasks can you actually use it for?

The tool focuses on read-oriented task management workflows, turning plain-language prompts into queries against the OmniFocus database via MCP. Use cases anchored in the feature set include asking "What should I do this afternoon?", listing stalled projects, checking availability by deferred starts, and scanning inbox items. The implementation emphasizes retrieval and analysis rather than full task creation, matching those who want conversational access to existing project and task metadata.

How accurate are the outputs compared to doing it manually?

Accuracy depends on the source database state and the AI client's interpretation. Project health detection and context-aware filtering report states such as active, on hold, or stalled directly from OmniFocus data, and time-based queries respect due dates and deferred starts. JXA integration reduces IPC delays, which supports near-instantaneous feedback, but users should validate AI-produced summaries against the app when decisions are important.

What file formats or inputs does it accept and what are its limits?

Inputs are the local OmniFocus database and MCP client prompts. The server requires macOS with OmniFocus installed and an MCP-compatible client, for example Claude Desktop, to send natural language queries. The implementation concentrates on querying and retrieval; task creation and modification are not the primary functions, though the MCP framework allows future expansion. Installation typically needs basic command-line edits to an MCP settings file.

Is it easy to set up and fit into existing workflows?

Setup suits technically inclined OmniFocus users who accept minimal configuration. The server runs locally on the Mac and uses JXA for high-performance communication with OmniFocus, which preserves local database access and supports the privacy-first architecture. The developer publishes the project openly and early adopters report reliable integration into MCP workflows. Note that the AI client you choose handles query processing and may operate off-machine.

A focused, technical solution for power users who want conversational OmniFocus access

The app is a practical fit for OmniFocus power users who use an MCP-capable AI client and are comfortable with light command-line configuration. Expect fast, local database querying and prepare to confirm AI suggestions against OmniFocus when decisions matter. Practical tip: enable descriptive prompts and test a few queries to calibrate the AI client's interpretation before relying on summaries for planning.