Moraya is an ultra-lightweight, privacy-first Markdown editor designed as a local AI agent platform. It empowers users with advanced AI capabilities directly on their machine, offering a seamless, secure, and extensible writing experience. Targeting developers, researchers, and power users, Moraya bridges cutting-edge AI with user sovereignty, ensuring data privacy and complete control.

How It Works

Built with Rust and Tauri v2, Moraya delivers native performance with a minimal ~10MB footprint and instant launch times. Its frontend utilizes Svelte 5 and a ProseMirror-based editor (Milkdown) for a true WYSIWYG Markdown experience. Core to its design is deep integration with local AI ecosystems via the Model Context Protocol (MCP), transforming the editor into a robust, self-hosted AI agent platform. This architecture prioritizes local operation and security, with all AI interactions processed on-device.

Quick Start & Requirements

Installation is straightforward across macOS, Windows, and Linux via direct downloads (DMG, MSI, DEB, AppImage) or Homebrew on macOS (brew install --cask moraya). Development requires Rust (stable), Node.js (>=18), and pnpm (v10.x). The application boasts a tiny memory footprint and instant launch.

Highlighted Details

• Multi-Provider AI: Supports numerous chat and image generation providers (OpenAI, Claude, Gemini, Ollama, etc.) with 71+ AI templates and streaming chat.
• MCP Ecosystem: Features a dynamic MCP container, one-click Marketplace for AI services, and support for autonomous agent workflows.
• Advanced Editor: Offers true WYSIWYG, full Markdown/GFM, KaTeX for math, Mermaid diagrams, syntax highlighting, and multi-window support.
• Security & Privacy: API keys stored in OS Keychain, all external calls proxied through a Rust backend, and full local operation ensures data never leaves the user's machine.
• AI Voice Transcription: Real-time speech-to-text with multiple providers, speaker diarization, and voiceprint recognition.

Maintenance & Community

The provided README does not detail specific contributors, sponsorships, or community channels (e.g., Discord, Slack).

Licensing & Compatibility

The README does not explicitly state the project's license. This omission requires clarification for understanding commercial use or derivative works.

Limitations & Caveats

macOS builds are not code-signed, necessitating a manual terminal command (xattr -cr) to bypass Gatekeeper warnings. PDF, DOCX, and EPUB export functionality is planned for future integration with pandoc.