jaylfc/taOS

Self-hosted AI agent OS. Your memory, chat, agents, and files stay on hardware you own, offline by default, cloud by choice. Offline AI memory (taOSmd), self-hosted multi-framework group chat, a full web desktop + app store, and auto-clustering across the consumer hardware you already have (Orange/Raspberry Pi, Mac mini, gaming PC).

What it solves

taOS is a self-hosted AI agent platform designed to give users full sovereignty over their AI's memory, data, and compute. It eliminates reliance on cloud providers by allowing users to run agents, manage long-term memory, and execute inference on their own hardware—ranging from old laptops and Raspberry Pis to gaming PCs and Android phones.

How it works

The system operates as a distributed compute cluster where a central controller manages multiple "worker" nodes. These workers can be any device with a compatible OS (Linux, macOS, Windows, Android), which the controller organizes into a compute mesh to handle LLMs, embeddings, and image/video/audio generation.

Key architectural components include:

  • Framework-Agnostic Core: taOS manages the agent lifecycle (memory, files, API keys) outside the specific agent framework. This allows users to switch between frameworks (e.g., SmolAgents, LangChain, OpenClaw) without losing history or configuration.
  • taOSmd Memory System: A specialized memory layer using a temporal knowledge graph and hybrid vector search to provide high-accuracy retrieval for agents.
  • Web Desktop Shell: A browser-based OS environment that provides a window manager and a suite of 40+ bundled apps for managing agents, models, and system health.
  • Skill Registry: A set of framework-agnostic tools (e.g., web search, code execution) that are translated via adapters to work across different agent frameworks.

Who it’s for

  • Privacy-conscious users who want their AI conversations and memory to remain offline and on their own hardware.
  • AI enthusiasts and developers who want to experiment with multiple agent frameworks and local models without data migration.
  • Home labbers looking to turn spare hardware (SBCs, old PCs) into a distributed AI compute cluster.

Highlights

  • Distributed Hardware Support: Supports Apple Silicon (MLX), NVIDIA, AMD, Rockchip NPU, and CPU-only devices.
  • Framework Independence: Switch agent frameworks while retaining all memory, files, and connections.
  • Comprehensive Web OS: Includes a full desktop experience with a window manager, app store, and mobile-optimized PWA mode.
  • High-Accuracy Memory: Features the taOSmd system with a temporal knowledge graph and contradiction detection.
  • Extensive Integration: Bundles 40+ apps, 109+ catalog apps, and 47 MCP plugins.