Voicetree offers a spatial Integrated Development Environment (IDE) designed for recursive multi-agent orchestration, addressing the challenges of managing numerous terminals and mitigating "context rot" in AI agent workflows. It provides a visual, Obsidian-style mindmap interface where agents, their tasks, planning documents, and progress updates are organized spatially, transforming projects into interactive graph views. This approach aims to enhance clarity, reduce cognitive load, and improve agent reliability for complex, long-term tasks.

How It Works

Voicetree employs a spatial computing paradigm, allowing users to arrange agents and their associated artifacts (like Markdown planning files and terminal outputs) in a visual graph. The core innovation lies in its transparent subagent orchestration, enabling complex tasks to be automatically decomposed into simpler, manageable subtasks that can run in parallel or sequence. Each subagent's activity is fully visible and controllable within the graph, unlike opaque processes in other tools. This design facilitates fractal project assembly from small, reliable agent outputs and leverages spatial proximity to denote relevance, creating an intuitive workflow.

Quick Start & Requirements

• Installation: Binaries are available for macOS (Apple Silicon/Intel) and Windows. Linux users can use the AppImage or an install script (curl -fsSL https://raw.githubusercontent.com/voicetreelab/voicetree/main/install.sh | sh). macOS users can also install via Homebrew (brew tap voicetreelab/voicetree && brew install voicetree).
• Development Prerequisites: Node.js 18+ and Python 3.13 are required, along with the uv Python package manager.
• Development Setup: Frontend requires npm install in frontend/webapp and npm run electron to start. Backend setup involves uv sync from the project root.
• Community: Join the Discord community at https://discord.gg/3z4Gbquv.

Highlighted Details

• Transparent Subagent Orchestration: Enables automatic decomposition of complex tasks into predictable, controllable subagents, with full visibility into their execution and outputs.
• Built for Flow: Maintains project state across sessions ("no cold starts"), offers spatial navigation where agents act as "landmarks," stores all data as local Markdown files for hackability, and improves LLM efficiency through context pruning.
• Voice & Thinking Tool: Supports a speech-to-graph mode for hands-free idea capture and serves as a spatial thinking tool, leveraging the graph structure for natural context compression.

Maintenance & Community

The project maintains a Discord community for support and feedback (https://discord.gg/3z4Gbquv). Direct contact for inquiries is available at hello@voicetree.io.

Licensing & Compatibility

The project is licensed under the Business Source License 1.1 (BSL 1.1), which converts to Apache 2.0 after four years. The BSL 1.1 may impose usage restrictions, particularly for commercial entities offering competing services, during the initial period.

Limitations & Caveats

The Business Source License 1.1 introduces potential restrictions on commercial use for the first four years, which may impact adoption for certain business models. Development requires specific versions of Node.js and Python, and the uv package manager. The README does not detail specific performance benchmarks or known bugs.