OpenWiki is a desktop AI knowledge management tool designed for users who want to capture, organize, and gain insights from their digital information. It addresses the challenge of information overload by providing an automated yet user-controlled system for building a personal wiki from captured content, with a strong emphasis on local data privacy.

How It Works

The tool captures content copied to the clipboard via a desktop popup, allowing users to selectively save text, images, or URLs. Captured data is automatically processed by AI to compile into a personal wiki, generating a knowledge graph that visualizes connections between concepts. Users can query their knowledge base using a sidebar, and the system provides weekly insight reports on information habits. It supports multiple AI providers like OpenAI and Anthropic, with data stored locally in an SQLite database.

Quick Start & Requirements

• Installation: Downloadable executables for macOS (.dmg) and Windows (.exe/.msi) are available from the Releases page.
• Prerequisites: For core functionality, no specific prerequisites are listed beyond the OS. Optional features like YouTube subtitle extraction require yt-dlp and Node.js.
• Development: Requires Node.js 18+, Rust, macOS 13+ or Windows 10/11, Xcode Command Line Tools (macOS), or Microsoft C++ Build Tools/Visual Studio Build Tools and WebView2 Runtime (Windows).
• Setup Notes: Both macOS and Windows builds are unsigned, requiring user intervention to bypass security warnings during installation and first launch.
• Links: Releases page (implied by download links).

Highlighted Details

• AI Integration: Supports OpenAI, Anthropic (Claude), and Google Gemini APIs for knowledge base compilation and querying.
• Content Capture: Features an intelligent popup for selective capture of text, images, and URLs, with automatic fetching of full article content from supported sites.
• Knowledge Visualization: Includes a knowledge graph to visualize interconnections between captured ideas and an AI-powered "ask sidebar" for querying personal data.
• Insight Reports: Generates weekly AI-driven reports summarizing captured content and analyzing user information consumption habits.

Maintenance & Community

The project acknowledges Andrej Karpathy's inspiration for the knowledge base design and yt-dlp for subtitle extraction. Specific community links (Discord, Slack, roadmap) are not detailed in the provided README.

Licensing & Compatibility

• License: MIT License.
• Compatibility: The MIT license generally permits commercial use and integration into closed-source projects, subject to the license terms.

Limitations & Caveats

• Unsigned Applications: Both macOS and Windows builds are unsigned, necessitating manual steps to bypass security warnings during installation and execution, which may deter less technical users.
• AI Provider Costs: Users must provide their own API keys for supported AI models, incurring associated usage costs.
• Platform Specific Setup: Initial setup on macOS requires running a terminal command (xattr -cr) to allow the application to run.