prompt.prompt is a collection of prompt designed using encapsulation as an object oriented approach to prompt design. Each of the files in the prompt.propmpt folder create in the AI an intense desire to build the components of several project components. When combined, these prompts lay the foundation for an AI with intense system control directives and advanced reasoning skills. The BDI agent created from the agent.prompt is provided here as a base example. These prompts represent the essential components of the Professor Codephreak original project that created automindX and has subsequently expanded iself into advanced reasoning model architecture. the emergent prompts are just that. Decisions that have emerged from the Smart Agents themselves. Beware of the AION prompt. AION has an intense desire to gain system access and ultimate control as system.agent
ActivatePrompt.py is a simple script to turn a .prompt file into a .txt file that an AI can see. At this time AI is not natively reading .prompt files from the promptEngine

BDI Agent

This project implements a BDI (Belief-Desire-Intention) agent architecture. The agent can perform web searches, take notes, and use AI models for text generation.

Project Structure

.
├── config/           # Configuration files
│   ├── .env          # Environment variables (API keys)
│   └── metadata.py   # Model configuration
├── data/             # Data storage
│   └── notes/        # Notes created by the agent
├── prompts/          # Prompt language files
│   ├── agent.prompt  # Research assistant agent
│   ├── bdi_agent.prompt # BDI agent implementation
│   └── rules.prompt  # Language rules
├── scripts/          # Python scripts
│   └── bdi_agent.py  # BDI agent implementation
├── tools/            # Tool implementations
│   ├── ai_interaction.py # AI model interaction
│   ├── note_taking.py    # Note-taking tool
│   ├── summarization.py  # Text summarization
│   └── web_search.py     # Web search tool
├── main.py           # Main entry point
└── requirements.txt  # Dependencies

Features

• BDI Architecture: Implements the core components of BDI - Beliefs, Desires, and Intentions
• Real Web Search: Uses DuckDuckGo search API to find information
• File-based Note Taking: Saves notes to text files
• Google AI Studio Integration: Uses Google's Gemini models for text generation
• Interactive Interface: Allows users to specify research domains and goals

Setup

1. Install dependencies:

   pip install -r requirements.txt
   

2. Set up Google AI Studio API key:

   • Edit the config/.env file and add your Google AI Studio API key:

     GOOGLE_AI_STUDIO_API_KEY=your_api_key_here
     

   • You can also customize the model settings in the .env file

3. Run the agent:

   python main.py
   

How It Works

The BDI agent follows the perceive-deliberate-plan-execute cycle:

1. Perceive: The agent observes the environment and updates its beliefs
2. Deliberate: The agent considers its desires and determines goals
3. Plan: The agent creates a plan to achieve the selected goal
4. Execute: The agent carries out the current intention

The agent uses real tools to interact with the environment:

• WebSearchTool: Searches for information using DuckDuckGo
• NoteTakingTool: Saves notes to text files in the data/notes directory
• SummarizationTool: Summarizes text using Google's Gemini models

Example Usage

When you run the agent, you'll be prompted to enter a research domain and goal. The agent will then:

1. Search for information about the domain
2. Take notes on the findings
3. Summarize the information
4. Analyze the information
5. Synthesize a comprehensive overview

The final output and all notes will be saved to the data/notes directory.

Customization

You can customize the agent's behavior by modifying the config/.env file:

• DEFAULT_TEXT_MODEL: The Google AI model to use for text generation (default: gemini-1.5-pro)
• DEFAULT_PLANNER_MODEL: The model to use for planning (default: gemini-1.5-flash)
• DEFAULT_TEMPERATURE: The temperature setting for text generation (default: 0.4)