A client-server system in Python that simulates a virtual agent for vehicle queries.
- MCP Protocol: Communication via Model Context Protocol.
- Vehicle Queries: Filter by brand, year, fuel type, etc.
- Conversational Agent: Interprets natural language inputs.
- Local Database: 500+ fictional vehicles with SQLite.
- Modular Architecture: Clean and scalable design.
- Python 3.10+
- SQLite + SQLAlchemy (ORM)
- Faker (test data generation)
tabulate(CLI table formatting)- TCP/IP Sockets
- Bash (automation scripts)
c2s/
├── client/ # Client-side agent
├── database/ # DB models, migrations, and seeding
├── server/ # MCP protocol server
├── main_client.py # Client entry point
├── main_server.py # Server entry point
├── setup.sh # 1-click setup script
├── requirements.txt # Dependencies
└── README.md # You’re here :)- Clone the repo::
To download the repository, run the command below.
git clone https://github.com/Victor5g/Agente-de-Veiculos-C2S.git
cd Agente-de-Veiculos-C2S- Set up the environment:
To set up the environment, install dependencies, and run the project, execute the command below in the root directory of the project under c2s/.
bash setup.shIf you encounter permission issues, run this first!!
chmod +x setup.shTip: On the first run, type y to accept the option to populate the database before performing the search!!!
🗄️ Deseja popular o banco de dados com dados randomicos ? (s/n): s