Proto-Mind

Product Introduction

1. Definition: Proto-Mind is a native macOS application (Apple Silicon, macOS 14+) that functions as a floating, persistent AI workspace. It technically integrates a multi-window interface, a local memory system, and a model-agnostic chat client into a single, cohesive desktop environment.
2. Core Value Proposition: It exists to eliminate context-switching and fragmentation in AI-assisted workflows. Its primary value is consolidating AI conversations, live web browsing, document editing, and project-specific memory into a single, always-available Mac workspace that can be minimized to an unobtrusive cube, enabling continuous parallel task execution.

Main Features

1. Floating, Collapsible Workspace: The core interface is a persistent, floating window group on the macOS desktop. Users can detach and resize individual companion windows (e.g., browser, document, secondary chat). The entire workspace can be "folded" into a small interactive cube that stays on screen while background tasks continue to run, providing immediate access without cluttering the desktop.
2. Model & Account Agnostic Parallel Chats: The software allows simultaneous, independent AI conversations. Each chat session can be configured to use a different AI model and account—including personal ChatGPT/Codex subscriptions, supported third-party APIs (like OpenAI), or local models via Ollama—giving users granular control over capabilities, costs, and context per task.
3. Project Memory & Voice Control: Proto-Mind features an editable, long-term memory system attached to specific projects, storing key decisions, preferences, and notes locally on the Mac. It also includes live voice control to initiate tasks, open projects, or interject corrections mid-task, using a separate OpenAI API key for speech-to-text processing.

Problems Solved

1. Pain Point: It addresses the fragmented nature of modern AI tool use, where users juggle between browser tabs for ChatGPT, separate windows for reference documents, and other apps for related work, leading to constant alt-tabbing, lost context, and disrupted flow states.
2. Target Audience: The primary user personas are macOS-based knowledge workers, including software developers, content creators, researchers, and product managers who regularly use AI for iterative tasks like coding, writing, research synthesis, and project planning and require a persistent, organized workspace.
3. Use Cases: Essential scenarios include: conducting research with an AI assistant while simultaneously viewing source websites and taking notes; iteratively drafting and editing a document based on AI feedback in a side-by-side window; managing multiple, concurrent AI-driven tasks (e.g., coding in one chat, writing copy in another) under different model accounts.

Unique Advantages

1. Differentiation: Unlike web-based ChatGPT or standalone desktop AI apps, Proto-Mind is not a single chat interface. It is a full desktop workspace manager designed for parallel, multi-context work. It differs from traditional window management tools by deeply integrating AI chats, local memory, and content windows into a unified, purpose-built system.
2. Key Innovation: The specific innovation is the "cube" metaphor—a persistent, minimal desktop presence that embodies the entire workspace state. This combines the power of a multi-monitor workflow with the simplicity of a single, collapsible object, allowing the AI workspace to be both omnipresent and non-intrusive, a concept unique among AI productivity tools.

Frequently Asked Questions (FAQ)

1. Is Proto-Mind a free AI tool? No, Proto-Mind itself is a free macOS application, but it acts as a workspace interface. To use AI models, you must supply your own accounts. It works with your paid ChatGPT Plus subscription (for Codex), your own OpenAI API keys (for voice and GPT API), or free, local models via Ollama.
2. How does Proto-Mind handle data privacy and where is data stored? Conversation history and project memory are stored locally on your Mac. Data is only sent to cloud providers (OpenAI, etc.) when you actively use those cloud-based models or the voice feature. Using local Ollama models keeps all processing entirely offline, enhancing privacy for sensitive projects.
3. Can Proto-Mind control my computer or automate tasks? It can work with files and execute commands on your Mac, but only after you explicitly enable "Mac access" permissions. For advanced screen control or automation, it requires the separate, signed "Computer Use" helper from Codex Desktop, which is not included with Proto-Mind and needs its own macOS permissions.
4. What are the system requirements for Proto-Mind? It requires an Apple Silicon Mac (M1, M2, M3, or newer) running macOS Sonoma (14.0) or later. It is currently an early beta and is not Apple-notarized, which requires specific steps to launch on first use, as detailed on its download page.
5. How does the project memory feature work? Project memory is a manual, editable note-taking system within the app. Users can save specific points from conversations (like design decisions or key requirements) to a project. These notes persist locally, can be reviewed, edited, or deleted, and are provided as context to the AI in subsequent conversations within that project.