Orxis
Orchestrate AI with Precision and Purpose.
Orxis is a structured AI instruction generator for developers, writers, and professionals who need clear, consistent control over how AI models behave. Choose between two modes: a guided step-by-step or Quick Mode โ a conversational AI that builds your complete system prompt in one shot. The output covers role definition, mission, a cognitive loop, context, boundaries, output formatting, and quick shortcuts โ ready to paste into ChatGPT, Claude, Gemini, or any other LLM. Available in 7 languages. No account required.
---
## ๐ Use Orxis Now
**No installation. No sign-up. Just open and build.**
> **[โ orxis.vercel.app](https://orxis.vercel.app)**
Orxis offers two modes โ choose whichever fits your workflow:
- **Step-by-Step** โ Walk through focused prompts about your persona, task, tone, and rules. Orxis assembles your Custom Instruction section by section.
- **Quick Mode** โ Describe what you need in plain language. The AI asks any clarifying questions it needs, then generates your complete instruction in one shot.
Copy the result and paste it directly into ChatGPT, Claude, Gemini, or any other AI assistant. No account, no subscription, nothing to install.
---
## โฆ Features
- **AI-Powered Instruction Generation** โ Both Step-by-Step and Quick Mode use Google Gemini to produce a structured, expert-level Custom Instruction precisely tailored to your context and goals.
- **Seven-Section Instruction Architecture** โ Role & Identity, Mission, Cognitive Loop, Context & Input, Boundaries & Execution Rules, Output Formatting, and Quick Shortcuts. Each section is guided by your inputs and produces a paste-ready system prompt.
- **Multilingual Support** โ The full interface, generated names, descriptions, conversation starters, knowledge suggestions, and quick shortcuts are all produced in the user's chosen language. Seven languages supported: English, Thai, Chinese, Japanese, Spanish, Korean, and Indonesian.
- **Multi-Model Integration Guides** โ Documentation for ChatGPT, Claude, and Gemini covering recommended injection methods and platform-specific behaviours for structured prompting.
- **Prompt Gallery** โ Browse community-contributed instruction templates at `/gallery`. Submit your own to share with other users.
- **Output Studio** โ Once your Custom Instruction is ready, Output Studio presents it formatted for each supported platform. Switch between Gems, GPTs, Projects, and Raw views โ copy any individual field or the entire result in one click, with platform-specific guidance on exactly where to paste it.
- **Progressive Web App** โ Installable directly from the browser on desktop and mobile.
- **Premium Dark UI** โ Glassmorphism surfaces, a cohesive neon accent palette, and smooth micro-animations throughout.
---
## โ๏ธ Tech Stack
| Layer | Technology |
|---|---|
| Framework | [Next.js](https://nextjs.org/) โ App Router, static prerendering |
| Language | [TypeScript](https://www.typescriptlang.org/) 5 |
| Styling | [Tailwind CSS](https://tailwindcss.com/) + CSS Custom Properties |
| Animation | [Framer Motion](https://www.framer.com/motion/) 12 |
| Icons | [Lucide React](https://lucide.dev/) |
| AI | [Google Gemini API](https://ai.google.dev/) |
| Deployment | [Vercel](https://vercel.com/) |
---
## ๐ Running Locally
### Prerequisites
- Node.js 18.17 or later
- npm, yarn, or pnpm
- A [Google Gemini API key](https://aistudio.google.com/apikey) (free tier available)
### Setup
```bash
git clone https://github.com/khotjarb/orxis.git
cd orxis
npm install
```
Create a `.env.local` file in the project root:
```env
# Required โ powers the AI instruction generator
GEMINI_API_KEY=your_gemini_api_key_here
```
Then start the development server:
```bash
npm run dev
```
Open [http://localhost:3000](http://localhost:3000) in your browser.
> **Without a Gemini API key**, the generator will still work using a local template fallback. The output will be more basic than the AI-generated version, but fully functional.
### Production build
```bash
npm run build
npm run start
```
---
## ๐ Architecture
- **Instruction generation** โ Your inputs from either Step-by-Step or Quick Mode are processed server-side to produce a complete seven-section Custom Instruction tailored to your context. Both modes share the same structured output format, ensuring consistent results regardless of which path you take.
- **Output Studio** โ Presents your generated instruction formatted for each supported platform (Gems, GPTs, Projects, Raw). Copy individual fields or the full result in one click, with step-by-step guidance on where to paste it in each AI platform.
- **Content pages** (documentation, changelog, legal) are statically prerendered at build time and served from the CDN edge โ no runtime compute required.
- **Community Gallery** โ Submissions are handled independently from the instruction generation flow, ensuring Gallery availability is never affected by generator load.
---
## ๐ Project Structure
```
src/
โโโ app/ # Next.js App Router โ pages and route layouts
โ โโโ api/ # Server-side API routes
โ โ โโโ generate/ # Step-by-Step AI instruction generation
โ โ โโโ quick-generate/ # Quick Mode conversational AI generation
โ โ โโโ gallery/ # Fetch community prompts
โ โ โโโ submit/ # Submit community prompts
โ โโโ (legal)/ # Privacy Policy, Terms of Service, License
โ โโโ about/ # About page
โ โโโ changelog/ # Release history
โ โโโ docs/ # Documentation (client-rendered)
โ โโโ gallery/ # Prompt Gallery
โ โโโ generate/ # Core instruction generator (Step-by-Step + Quick Mode)
โ โโโ pricing/ # Pricing page
โ โโโ settings/ # User settings
โ โโโ support/ # Support page
โ โโโ use-cases/ # Use Cases showcase
โ โโโ layout.tsx # Root layout โ global metadata
โโโ components/ # Shared UI components
โ โโโ docs/ # Documentation sub-components and per-model guides
โโโ data/ # Static data files (domains, presets)
โโโ i18n/ # Internationalisation โ LanguageProvider, useT hook
โ โโโ locales/ # Locale JSON files for all 7 supported languages
โโโ globals.css # Design token system โ CSS custom properties
```
---
## ๐ค Contributing
Issues and pull requests are welcome. For significant changes, please open an issue first to discuss what you would like to change.
---
## ๐ License
This project is **not open-source for commercial use**. It is released under a custom **Proprietary Non-Commercial License** โ you may clone and run it locally for personal or educational purposes, but commercial deployment, white-labeling, and redistribution are strictly prohibited. See the [`LICENSE`](./LICENSE) file for full details.
---