Discover and explore top open-source AI tools and projects—updated daily.
milind-soniAI-powered macOS assistant for screen and voice control
Top 70.2% on SourcePulse
TipTour offers an open-source, AI-powered macOS companion for voice and screen control. It enables users to point, click, type, open apps, edit text, or act on highlighted screen areas via natural language, enhancing productivity.
How It Works
TipTour integrates Gemini Live for real-time voice, screen understanding, and tool calling. It uses CUA Driver Core for computer control (clicks, typing) and macOS Accessibility for native app structure analysis. A unique "Focus Highlight" feature lets users paint a freeform screen area (Ctrl+Shift+drag) for context-specific commands. This enables precise, AI-driven computer manipulation.
Quick Start & Requirements
Building from source requires macOS 14+, Xcode 16+, and Node 20+ (for optional Cloudflare Worker). Setup involves opening tiptour-macos.xcodeproj in Xcode, selecting a signing team, building (Cmd+R), pasting a Gemini API key into the menu bar panel, and granting macOS permissions (Microphone, Screen Recording, Accessibility, Screen Content). The API key is stored in macOS Keychain.
Highlighted Details
Maintenance & Community
The provided README does not contain specific details regarding notable contributors, sponsorships, community channels (like Discord or Slack), roadmaps, or deprecation notices.
Licensing & Compatibility
The project is licensed under the MIT license, which generally permits commercial use and integration into closed-source projects. Compatibility is restricted to macOS 14 and later.
Limitations & Caveats
TipTour is macOS-only (14+). Users must supply their own Gemini API key, incurring potential costs and privacy concerns. Building requires specific Xcode/Node.js versions. Potential TCC permission state invalidation exists when using xcodebuild for testing.
1 week ago
Inactive
elfvingralf
sohzm
OthersideAI