Hearth
Incubator Controls Projects
Jump Into App
Shortcut-first execution layer for builders and managers

The shortcut-first todo list that actually compounds.

Hearthflow is a beautiful Eisenhower-native command center: capture fast, prioritize hard, and execute with ruthless clarity. Behind the board, an AI-first incubator transforms raw ideas into actionable work, while built-in MCP connectivity lets you push context into Hearth from Cursor, Claude, Codex, and any LLM chat.

productivity flywheel signal-first planning AI-native execution context compounding high-agency workflow
Jump Into App See architecture
Shortcut-first keyboard actions: C, I, P, Cmd+K
Eisenhower-native Do / Delegate / Schedule / Drop
AI incubator Raw -> Shaped -> Bet -> promote to task
MCP-ready core Send from any LLM chat into Hearth in one step

Why It Wins

Hearthflow is not another pretty checklist. It is a full-stack execution system engineered for velocity: a shortcut-first interface, an Eisenhower operating model, and an AI-first incubator wrapped around a daemon + backend architecture that keeps every loop connected.

React Frontend

Speed surface, not admin panel

  • Kanban board with drag-and-drop and status flow
  • Keyboard-first task ops and command palette execution
  • Eisenhower matrix priorities baked into daily triage
  • Projects, archive, and scoped work/personal filtering
  • Control runner UI for Claude, Codex, and Bash actions
AI-First Incubator

Raw ideas -> production work

  • Capture fuzzy ideas instantly and stage them by confidence
  • Move from Raw to Shaped to Bet with minimal friction
  • Use MCP to send insights from Cursor, Claude, or Codex straight into Incubator
  • Promote matured ideas straight into executable tasks
  • Parse AI output into structured suggestions automatically
  • Turn strategy drift into an execution flywheel
Daemon + Backend + MCP

Operational backbone

  • Local Rust daemon runs your coding assistant inside Hearth and streams live updates
  • Memory file system for persistent context and daily logs
  • Kotlin Ktor backend with Neon JWT auth and user isolation
  • Task, idea, project, and sync routes with shared API surface
  • MCP server support to ingest context from external LLM chats into the same workspace
Key claim: Hearthflow compresses planning, prioritization, and AI-assisted execution into one continuous high-signal loop.

The Compound Execution Loop

A four-step engine designed to maximize throughput without sacrificing strategic focus.

01 Capture

Grab tasks and ideas at keyboard speed while context is still fresh.

02 Prioritize

Use the Eisenhower matrix to force hard focus decisions instantly.

03 Incubate

Run AI controls to shape raw concepts into clear next actions.

04 Execute

Ship, archive, and feed learning back into persistent memory.