A lightweight, browser-first tool for designing well-structured AI prompts. It offers a clean UI inspired by Claude.ai, live previews in multiple formats, a local prompt library, and direct testing with over 100 AI models via OpenRouter integration. The tool benefits users by simplifying prompt engineering with organized input sections, dynamic lists for constraints and examples, and real-time feedback, all while prioritizing privacy and local data storage.

How It Works

This project provides a modern, browser-based interface for crafting AI prompts. Core functionality includes structured input fields for Title, Role, Task, and Context, alongside dynamic lists for Constraints and Examples. A key advantage is the live preview feature, which renders the prompt in JSON, Markdown, and Plain Text formats in real-time, allowing for immediate edits. For AI interaction, it integrates with OpenRouter, enabling testing against numerous models with adjustable parameters like temperature, max tokens, and sampling methods, all processed client-side without a backend.

Quick Start & Requirements

1. Clone the repository: git clone https://github.com/Siddhesh2377/structured-prompt-builder.git
2. Navigate to the directory: cd structured-prompt-builder
3. Serve locally: npx serve . or simply open index.html in your browser.
4. Prerequisites: An OpenRouter API Key is required for AI testing. This key is stored securely in the browser's localStorage and is only sent to OpenRouter.

Highlighted Details

• Live preview and direct editing of prompts in JSON, Markdown, and Plain Text formats.
• Persistent local prompt library using browser localStorage, requiring no account or internet connection for saving.
• Seamless integration with OpenRouter for testing against 100+ AI models, with dynamic model loading and advanced parameter configuration.
• Five distinct, aesthetically pleasing themes (Default, Dark Slate, Dark Midnight, Light Warm, Light Cool) with persistent user selection.
• Strictly browser-based architecture with no backend, ensuring privacy and security by keeping all data and API keys local.

Maintenance & Community

Contributions are welcomed via bug reports, feature suggestions in discussions, and pull requests. The project maintains a roadmap with planned enhancements including prompt export/import, versioning, collaborative sharing, and a template marketplace. Development guidelines emphasize a "no-build" philosophy using vanilla JS/React UMD, maintaining a clean aesthetic, and ensuring mobile responsiveness. Sponsorship is encouraged to support ongoing development and new features.

Licensing & Compatibility

The project is released under the MIT License, allowing for broad use and modification. Its entirely client-side nature and use of local storage make it highly compatible with closed-source applications and environments where data privacy is paramount, as no data leaves the user's browser except for direct API calls to OpenRouter.

Limitations & Caveats

The project is actively under development, with several features listed on the roadmap, such as prompt versioning, export/import functionality, and collaborative sharing, not yet implemented. AI model testing is dependent on the availability and API of OpenRouter. The "no-build" philosophy, while simplifying setup, might present challenges for users requiring deep integration with complex build systems or custom extensions beyond vanilla JavaScript capabilities.