Focus Relay

Inspiration

An interruption can erase the small detail needed to restart. Focus Relay treats that detail as explicit state rather than relying on memory.

What it does

It adds a manually decomposed task with minutes and context, returns active work before queued work, pauses only active work with a resume note, and completes active work while retaining event history.

How we built it

Python domain tools persist a single-user queue in SQLite and are exposed through a local Python MCP server tested via Streamable HTTP (MCP 2025-11-25). A static browser replay, when used, is a JavaScript/localStorage equivalent and is not connected to MCP or Alexa.

Challenges

The useful behavior is in the transitions: paused context must be returned before selecting new work, and invalid task-state changes must be rejected.

What we learned

A narrowly scoped next action plus a resume note can be more reliable than pretending to automate task decomposition.

Accomplishments

The local workflow and fresh Streamable HTTP transcript verify adding, activating, pausing, and resuming task context.

What's next

Add multi-user isolation, calendar integration, richer history views, and reviewed background reminders only with explicit permissions.

This is a local single-user prototype: no cloud or LLM claim, no calendar, alarm, background job, notification, Alexa/Echo integration, or automatic decomposition.

Built With

• css3
• fastmcp
• html5
• javascript
• json-schema
• localstorage
• model-context-protocol
• pytest
• python
• sqlite
• streamable-http