AnonymousThinker is a premium AI reasoning platform designed to provide structured, logical responses with a specialized focus on defending and explaining Islamic principles through reason and scholarly knowledge.
| Feature | Description |
|---|---|
| 📌 Pin Conversations | Pin important conversations to the top of the sidebar |
| 🗄️ Archive Conversations | Archive chats instead of deleting them (right-click menu) |
| 🔍 Sidebar Search | Search conversations by title or content |
| 👍👎 Message Reactions | Rate AI responses as helpful or not helpful |
| 📤 Export Conversation | Download full conversation as a Markdown file |
| 📊 Word Count | AI responses show word count for quick reference |
| 🔢 Character Counter | Live character count in input (max 4000 chars) |
| 🖱️ Right-Click Context Menu | Right-click any conversation for rename, pin, archive, export, delete |
| 📊 Response Counter | Topbar shows number of AI responses in current chat |
| 🚀 Improved Error Toast | Cleaner error display, auto-dismisses after 5s |
- Dual-Model Comparative Analysis: Compare responses from a personal logic model (Llama 3.1 8B) and a powerful reasoning model (DeepSeek-R1-70B) side-by-side.
- Sovereign AI Hub: A centralized command center for administrators to define the AI's persona, logical framework, and core strategy.
- Knowledge Base (RAG): Ground AI responses in scholarly Islamic knowledge by uploading PDFs and text files globally.
- Admin-Restricted Training: Only authorized developers can train the AI or modify its core directives.
- Glassmorphic UI: A premium, modern interface with dark mode, fluid animations, and a focus on readability.
- Frontend: React.js, React Router, Axios, Vanilla CSS (Standardized Design System).
- Backend: Node.js, Express, MongoDB (Mongoose), JWT Auth.
- AI Infrastructure:
- Inference: Groq API & Hugging Face Serverless Inference.
- Reasoning Models: Llama-3.1-8B-Instruct & Llama-3.3-70B-Versatile.
- Logic Engine: Custom NLP pipeline for PDF parsing and RAG (Retrieval Augmented Generation).
- Node.js (v18+)
- npm or yarn
- MongoDB (Local instance or Atlas URI)
git clone <repository-url>
cd AnonymousThinkercd backend
npm install
cp .env.example .env
# Edit .env with your credentials
npm run devcd ../frontend
npm install
npm startnode scripts/makeAdmin.js your-email@example.comConfigure ADMIN_EMAIL and ADMIN_PASSWORD in your .env.
Right-click any conversation in the sidebar to access:
- Rename — Edit the conversation title inline
- Pin to top / Unpin — Keep important conversations accessible
- Export as .md — Download the full conversation as a Markdown file
- Archive — Hide the conversation without deleting it
- Delete — Permanently remove the conversation
Each AI response now shows thumbs up/down buttons. Click to rate:
- 👍 Helpful — Marks the response as useful
- 👎 Not helpful — Marks the response as needing improvement
Reactions are stored per-message in the database and can be used to track AI quality over time.
Click the search icon (🔍) in the sidebar header to activate search. Type to filter conversations by:
- Conversation title
- Last message preview
- Login as Admin
- Click "Train AI" in the sidebar
- Upload PDFs in the Upload Zone
- Update the "Core Logic & Strategy" textarea
- Click "Commit Changes to Core"
MIT License - Developed for Sovereign Thought Analysis.
"I am AnonymousThinker, an AI to understand different thoughts and to create a conclusion from them."