OpenPencil
The world's first AI-native open source vector design tool supports concurrent Agent teams, design-as-code (.op JSON) and multimodal intelligence.
###Basic operations 1. ** Create a design **: Click on the blank space of the canvas and use the toolbar (R= rectangle, O= ellipse, T= text, P= pen, L= straight line) to draw 2. **AI design generation **: Press `Cmd+J` to open the AI chat panel and enter a natural language description to generate the UI 3. ** Export code **: Press `Cmd+Shift+C` to open the code panel and select the target framework (React/Vue/HTML, etc.) 4. ** Save file **: Press `Cmd+S` to save as `.op` file (JSON format, supports Git version management) ##Functional characteristics - ** Prompt words → Canvas *: Describe the UI in natural language and generate it on infinite canvases in real time with streaming animation - ** Concurrent Agent team **: The orchestrator decomposes complex pages, and multiple AI agents process different blocks in parallel - ** Multi-model intelligence **: Automatically adapts to Claude / GPT-4o/ Gemini / DeepSeek / Qwen and other models - ** Design is code **: The `.op` file is JSON and supports Git, diff, code export (React/Vue/Flutter, etc.) - ** Built-in MCP server **: Install it into Claude Code / Codex / Gemini / OpenCode / Copilot CLI with one click - ˇ* * Complete drawing tools **: Rectangle, ellipse, pen, polygon, Frame, text, Boolean operation - ˇ **Figma import **: Support `.fig` file import - ** Design system **: Design variables, multiple themes, component systems
