⚡ KIRA - Knowledge Integration & Response Agent v1.3

Developed by ALAN CYRIL SUNNY
If you like this project, please ⭐ star the repository!

---

🧠 KIRA - Knowledge Integration & Response Agent

A fast, elegant, and responsive terminal-based AI assistant powered by your local liquid/lfm2-1.2b model via LM Studio.

• 💬 Real-time, token-by-token streamed output
• 🎨 Beautiful, colorful terminal UI with rich and ASCII banners via pyfiglet
• 🧠 Multi-turn conversation memory
• 🧬 Identity-aware (Name: KIRA, Version: 1.3, Creator: Alan Cyril Sunny)
• ⌨️ Glitchy effects and slow typing for hacker vibes
• ⚙️ Fully configurable API/model settings
• 🔒 100% local processing for privacy and control
• ⚡ Supports both NVIDIA CUDA and AMD ROCm GPUs for accelerated inference

---

✨ Features

• Identity Awareness: KIRA always knows her name, version, and creator.
• Streamed Output: See responses appear live, token by token.
• Rich Terminal UI: Styled with rich and ASCII art banners.
• Conversation Memory: Remembers previous turns for context.
• Customizable: Easily tweak API/model via environment variables.
• Local & Private: All data stays on your machine.
• GPU Support: Works with NVIDIA CUDA and AMD ROCm supported GPUs.

---

🛠️ Tech Stack

• Language: Python 3.8+
• Terminal UI: rich, pyfiglet
• AI Model: liquid/lfm2-1.2b (via LM Studio)
• Model Serving: LM Studio (REST API)
• API Communication: REST API (HTTP)
• GPU Acceleration: NVIDIA RTX GPU (CUDA) or AMD GPU (ROCm) for faster inference

---

💻 Requirements

• Python 3.8 or higher
• LM Studio running locally at http://localhost:1234
• Installed model: liquid/lfm2-1.2b or compatible
• (Optional) NVIDIA CUDA or AMD ROCm supported GPU for acceleration

---

🚀 Installation

1. (Optional) Create and activate a virtual environment:

   python -m venv .venv
   .venv\Scripts\activate   # Windows

2. Install the required Python packages:

   pip install requests rich pyfiglet

3. Start LM Studio with the liquid/lfm2-1.2b model and keep it listening on http://localhost:1234.

---

⚡ Quick Start

With LM Studio running, launch KIRA:

python chatbot.py

You'll see the stylized ASCII banner and can chat live in your terminal.
Type 'exit' to gracefully disconnect.

---

📝 Usage

1. Run the bot in your terminal.
2. Chat with KIRA in natural language.
3. Enjoy real-time, styled responses with memory and glitch effects.

---

💡 Example Prompts

• "Who are you?"
• "Summarize the latest conversation."
• "Tell me a programming joke."
• "What's your version and creator?"

---

🔧 Environment Variables (Optional)

Variable	Description	Default
LM_STUDIO_API_URL	LM Studio API endpoint	http://localhost:1234/v1
MODEL_NAME	Model name used for completions	liquid/lfm2-1.2b

---

📸 Demo

---

📁 Project Structure

📦 K.I.R.A/
 ┣ chatbot.py                # Main chatbot script
 ┗ README.md              # Project README

---

🧠 Identity Prompt

KIRA is always aware of her identity:

Your name is KIRA.
Your version is 1.3.
Your creator is Alan Cyril Sunny.
You are based on the model liquid/lfm2-1.2b.

---

👨‍💻 Developer

Alan Cyril Sunny
📫 alancyrilsunny@protonmail.com

---

📜 License

MIT License. Free to use, modify, and share.

---

✨ “May the shadows keep you safe.” — KIRA