Synthetic Autonomic Mind

Synthetic Autonomic Mind is an open-source AI ecosystem designed for macOS and Linux, offering a suite of tools focused on privacy and local data control. The platform consists of three main components: SAM, CLIO, and ALICE, each targeting distinct aspects of AI-assisted productivity and creativity. All tools are built to operate offline with local models or can connect to cloud providers when chosen by the user, ensuring that data remains private unless explicitly shared.

SAM is a native macOS application serving as an AI assistant that helps with daily tasks such as research, document drafting, trip planning, and budget management. It features a conversational interface, supports voice commands, and can remember previous conversations and user preferences through semantic memory and mini-prompts. SAM is capable of running completely offline using local MLX or GGUF models, but can also connect to external AI services like OpenAI, GitHub Copilot, and Google Gemini. It allows users to import and semantically index documents—including PDFs, Word files, text, and images—for information retrieval and referencing within conversations. SAM can be accessed from any device on a network via SAM Web.

CLIO is a terminal-native AI development agent that reads, modifies, and tests code, then commits changes and iterates as needed. It is designed to work locally on a range of devices, from Raspberry Pi to M4 Mac, and can operate over SSH or across multiple systems in parallel. CLIO remembers context across sessions, enabling seamless continuation of work, and does not require additional dependencies. It is suited for tasks such as code refactoring, code generation, and problem-solving directly from the terminal.

ALICE is an image generation tool that runs Stable Diffusion models locally on macOS or Linux. It supports text-to-image, image-to-image, LoRA, and provides a private gallery, a real-time GPU dashboard, and a built-in model marketplace for CivitAI and HuggingFace. ALICE integrates with SAM or can function independently via a web interface or API, with no subscription or per-image costs.

0. Installation is straightforward, requiring no account and featuring no telemetry. Synthetic Autonomic Mind emphasizes privacy, extensibility, and a methodology for effective human-AI collaboration called the Unbroken Method, which focuses on continuous context, ownership, and learning from documented failures.