A modern, low-code visual programming IDE for algorithm and AI workflow development, CanvasMind offers drag-and-drop orchestration, dynamic expressions, and embedded Python scripting, enabling seamless transitions from development to deployment. Its advanced features, including complex control flow, remote execution, and LLM integration, provide significant advantages over traditional tools for building sophisticated AI pipelines.

How It Works

Built on NodeGraphQt and qfluentwidgets, CanvasMind uses a visual, node-based paradigm. Its core innovation is dynamic parameter binding via expressions and global variables, supporting full conditional branching, iteration, and loops, unlike static, serial-only tools. It uniquely allows embedded Python for custom logic, one-click export to standalone runnable projects (API, CLI, Docker), and hybrid runtime orchestration (local IPython, remote SSH). Deep LLM integration provides intelligent code completion and agent-based assistance.

Quick Start & Requirements

Requires Python 3.8+ and PyQt5 or PySide6. Installation involves cloning the repository, creating a virtual environment, and installing dependencies:

git clone https://github.com/martin98-afk/CanvasMind.git
cd CanvasMind
python -m venv .venv
source .venv/bin/activate  # Linux/Mac
.venv\Scripts\activate     # Windows
pip install -r requirements.txt

Run with python main.py. Documentation: https://canvasmind-sphinx-build.readthedocs.io/zh-cn/latest/. Demo: https://www.bilibili.com/video/BV153zCBGEU2.

Highlighted Details

• Dynamic UI & Data: Complex form/tree controls with context-aware validation and dynamic property grids.
• Distributed Execution: Parallel DAG execution, hybrid runtime (local IPython, remote SSH), and selective in-memory caching.
• LLM Assistant: "LLM-Chatter" with multi-agent support, 30+ tools, and persistent long-term memory.
• Standalone Export: One-click export of workflows/subgraphs into self-contained, zero-dependency runnable projects.
• Plugin Triggers: Dynamic loading of Cron, Webhook, File Watchers with auto-adaptive UI.

Maintenance & Community

Active issue tracking and discussions on GitHub. Roadmap includes "In Progress" code-to-canvas auto-creation and "Planned" cloud execution. Contributions are welcomed.

Licensing & Compatibility

Licensed under the GPLv3 License. This copyleft license may restrict integration with proprietary software, requiring careful review for commercial adoption.

Limitations & Caveats

Cloud execution support is "Planned," not implemented. "Code-to-canvas auto-creation" is "In Progress." The GPLv3 license necessitates careful consideration for commercial use and closed-source integration.