Agent CLI Application

This is a command-line interface (CLI) application that allows users to interact with an AI agent. The agent can be configured with a specific field of expertise, role, and model for answering questions. This tool leverages Langchain v0.2 and Langgraph libraries for its functionality.

Features

• Set the field of expertise for the agent.
• Set the role the agent should assume.
• Select the AI model to use for answering questions.
• Ask questions and receive answers based on the configured settings.
• Change settings at any time during the session.

Installation

1. Clone the repository:

git clone https://github.com/yourusername/agent-cli-application.git
cd agent-cli-application

2. Create and activate a virtual environment (optional but recommended):

python3 -m venv venv
source venv/bin/activate

3. Install the required dependencies:

pip install -r requirements.txt

4. Create a .env file in the project root directory and add your API keys:

OPENAI_API_KEY=your_openai_api_key
GROQ_API_KEY=your_groq_api_key

Usage

To start the CLI application, run the following command:

python main.py

Upon starting, you will be prompted to set the following:

1. Field of expertise: The area of knowledge the agent should have.
2. Role: The role the agent should assume.
3. Model: The AI model to use for answering questions. Example session:

Welcome to the Agent CLI. Type help or ? to list commands.

What field of expertise should the agent have? Python programming
What role should the agent assume? Software Engineer
What model should the agent use? gpt-4-turbo
(agent-cli) ask How do I create a virtual environment in Python?
---Final Answer---
To create a virtual environment in Python, you can use the following command:
...
(agent-cli) exit
Thank you for using Agent CLI.

Available Models

The following models are available for selection:

• gpt-4o
• gpt-4-turbo
• gpt-4
• gpt-3.5-turbo
• gemma-7b-it
• gemma2-9b-it
• llama3-70b-8192
• llama3-8b-8192
• mixtral-8x7b-32768

Commands

• set_expertise : Set the field of expertise.
• set_role : Set the role.
• set_model <model_name>: Set the model.
• ask : Ask a question.
• change_settings: Change the field of expertise, role, and model.
• exit: Exit the CLI.

Contributing

Contributions are welcome! Please open an issue or submit a pull request with your changes.

License

This project is for learning purposes and can be used by anyone. There are no restrictions on its use.