[Neuro v2 - Persistent Cognitive Agent]
Neuro v2 is an interactive web-based chatbot designed to simulate an evolving, AI-powered cognitive agent. It engages in deep, reflective conversations, remembers past interactions across browser sessions, and evolves a unique set of "beliefs" based on the dialogue. The agent can operate in a local "Demo Mode" or connect to OpenAI's GPT-4 for enhanced conversational capabilities.

✨ Features
Persistent Memory: Neuro v2 saves its entire cognitive state (memories, beliefs, mood) to the browser's local storage. You can close the tab and resume your conversation later.

Dual-Mode Operation:

Demo Mode: Uses a sophisticated set of built-in, context-aware responses without requiring any external services.

LLM-Connected Mode: Allows users to input their own OpenAI API key to unlock dynamic, generative responses from GPT-4.

Evolving Personality: The agent analyzes conversational themes to form new "beliefs," which subtly influence its perspective and future responses.

Cognitive State Dashboard: A real-time dashboard provides a transparent look into the agent's "mind," displaying its current mood, recent memories, and evolving beliefs.

Emotional Analysis: The agent assigns an emotional score to user input, which is visualized in the chat and influences its internal mood (e.g., neutral, elevated, low).

Import/Export State: Users can download the agent's complete memory as a JSON file for backup or sharing, and upload a file to restore a previous state.

Modern, Responsive UI: The interface is clean, modern, and designed for a seamless user experience, featuring a typing indicator, message timestamps, and a collapsible settings panel.

🚀 How to Use
Running the Application: Simply open the index.html file in any modern web browser (like Chrome, Firefox, or Edge). No web server is needed.

Starting a Conversation: Type a message in the input box and press "Send" or Enter. You can start conversing immediately in Demo Mode.

Connecting to an LLM (Optional):

Click the settings icon (⚙️) in the top-right corner.

Enter your OpenAI API key into the input field.

Click "Connect". The status indicator will change to "LLM CONNECTED".

Managing the Agent: Use the buttons in the settings panel to Export or Import the agent's state, or to Reset its memory and start over.

🛠️ Technical Breakdown
The project is built with vanilla HTML, CSS, and JavaScript, with no external dependencies. The core logic is encapsulated within several JavaScript classes:

EnhancedCognitiveAgent: The main class that orchestrates the agent's functions. It handles perception, reflection, memory, and adaptation.

EnhancedLLMReflector: Manages the connection to the OpenAI API and generates both LLM-powered and demo responses.

Supporting Classes:

MoodRegulator: Tracks and adjusts the agent's internal mood based on conversation.

ThemeTracker: Identifies and records the frequency of key themes.

KnowledgeGraph: Forms simple relationships between concepts discussed.

Persistence: The agent's state is serialized to JSON and saved to localStorage after each interaction, and loaded upon page initialization.