PromptStack is a modern AI prompt library and playground built for developers, prompt engineers, content creators, and product teams. It gives you a centralized workspace to create, organize, test, and reuse prompts without switching between notes, chats, or scattered text files.
Instead of keeping prompts in random places, PromptStack helps you manage them like a real product: store templates, inject variables, choose model settings, run them live, and refine them quickly.
PromptStack is useful when you want to:
- Keep all your prompt ideas and reusable templates in one place
- Build prompts faster with structured templates and placeholders
- Test prompts live with real AI providers before using them in production
- Reuse successful prompts across different projects and workflows
- Organize prompts by category, tags, and purpose
- Compare outputs from different providers and model configurations
PromptStack provides a complete workflow for prompt development:
- Create and save reusable prompt templates
- Use dynamic placeholders such as
{{topic}},{{audience}}, or{{tone}} - Search and filter prompts by category, title, or keyword
- Run prompts directly in the browser using Gemini or Groq
- Add system instructions and model settings for each prompt
- Store prompts in MongoDB for fast reuse and iteration
- Test and refine prompts without switching between multiple tools
- Browse a curated template marketplace and fork templates into your library
- Track workspace-level stats such as prompt count, template count, runs, and provider usage
- Prompt library dashboard with search, filtering, and live loading states
- Reusable prompt templates with dynamic variable support
- Interactive playground for executing prompts with user-provided values
- Template marketplace with featured templates and fork support
- Prompt detail pages with full template preview and integration snippets
- Workspace analytics for total prompts, templates, runs, and providers
- Responsive UI built with Next.js, Tailwind CSS, and shadcn/ui
- Theme toggle and polished modern interface
- MongoDB-backed persistence for prompts, templates, and stats
- Home page: browse and manage your prompt library
- Templates page: explore public or curated prompt templates
- Help page: understand how to use PromptStack efficiently
- Prompt detail page: inspect prompt content, variables, and execution details
- Template detail page: view template metadata, usage stats, and code examples
- Playground page: execute prompts live with dynamic inputs
/— Prompt library dashboard/templates— Template marketplace/help— Documentation and usage guide/:id/prompt— Prompt detail page/:id/template— Template detail page/:id/playground— Live prompt playground
PromptStack is built with a modern full-stack setup:
- Next.js 16
- React 19
- TypeScript
- Tailwind CSS
- shadcn/ui
- MongoDB + Mongoose
- Gemini API integration via
@google/genai - Groq API integration via
groq-sdk - Sonner for toast notifications
src/app/— app routes, page components, and server actionssrc/components/— reusable UI components and dialogssrc/global/— shared types, constants, and app-wide configsrc/lib/— database connection, parsers, helpers, and utilitiessrc/models/— Mongoose schemas for prompts and templatespublic/— static assets and screenshots
Make sure you have:
- Node.js 20 or newer
- A MongoDB instance or MongoDB Atlas cluster
- A Gemini API key and/or a Groq API key
Create a .env.local file in the project root and add:
MONGODB_URI=your_mongodb_connection_string
GEMINI_API_KEY=your_gemini_api_key
GROQ_API_KEY=your_groq_api_keynpm installnpm run devThen open http://localhost:3000 in your browser.
npm run dev— start the development servernpm run build— create a production buildnpm run start— run the production build locallynpm run lint— run ESLint checksnpm run format— format the project with Prettier
PromptStack is a practical and polished solution for anyone who works with AI prompts regularly. It combines prompt storage, variable-based templating, live execution, provider support, and template sharing in one clean interface.
