Diagen is a command-line tool and library designed to automatically generate visually appealing diagrams from textual data. It targets developers and technical users who need to quickly visualize complex information, leveraging multiple AI models for creation and iterative refinement, thereby streamlining the diagramming process.

How It Works

Diagen utilizes a multi-stage AI approach. Users provide source data and descriptions via an interactive CLI or a programmatic API. The tool then employs specified AI models (OpenAI, Anthropic, Google) for initial diagram generation, followed by refinement steps involving AI-driven critique and self-correction. This iterative reflection process aims to improve diagram clarity and accuracy automatically.

Quick Start & Requirements

• Primary install/run command: npx diagen <source file> or npm install diagen for programmatic use.
• Non-default prerequisites: Node.js (v14 or later), d2 (Diagram rendering tool). Requires API keys for OpenAI, Anthropic, or Google (Gemini) set as environment variables.
• Setup involves installing Node.js and d2, cloning the repository (optional), installing dependencies (bun install or npm install), and configuring API keys.
• Links to official quick-start, docs, or demo pages are not explicitly provided in the README.

Highlighted Details

• Supports diverse diagram types: flowcharts, architecture diagrams, entity-relationship diagrams, process flow diagrams, and mind maps.
• Flexibility in choosing AI models for generation, fixing, and critique stages.
• Features an automatic iterative refinement loop for diagram improvement based on AI feedback.
• Offers both a user-friendly CLI interface and a programmatic TypeScript API.

Maintenance & Community

The project is created by Hrishi Olickel and welcomes community contributions via standard pull requests. No specific community channels (e.g., Discord, Slack) or roadmap links are mentioned in the README.

Licensing & Compatibility

Licensed under the Apache 2.0 License. This permissive license generally allows for commercial use and integration into closed-source projects, provided standard attribution and license terms are met.

Limitations & Caveats

The README does not provide performance benchmarks or detailed setup time estimates. Specific hardware requirements beyond Node.js and d2 are not detailed, and users are responsible for managing their own AI model API costs and rate limits. The effectiveness of the automated diagram refinement heavily relies on the chosen AI models and the quality of the input data.