Logo Lanfrica

Ahmzyjazzy/naija-tax-agent

Domaine:

socioeconomic

Type de record:

software
Créateur:
Ahm
Hôte:
Naija Tax Agent is an intelligent agent that provides assistance with Nigerian tax-related queries and business registration matters. The agent leverages Google's Gemini 2.0 model and includes specialized tools for searching business registrations through Nigeria's Corporate Affairs Commission (CAC). # Naija Tax Agent Naija Tax Agent is an intelligent agent that provides assistance with Nigerian tax-related queries and business registration matters. The agent leverages Google's Gemini 2.0 model and includes specialized tools for searching business registrations through Nigeria's Corporate Affairs Commission (CAC). ## Features - 🤖 AI-powered responses using Google's Gemini 2.0 model - 🔍 CAC public search integration for business name verification - 🌐 Google search grounding for up-to-date tax information - 📚 Expert knowledge in Nigerian tax matters ## Prerequisites - Python 3.11 or higher - Google ADK account and credentials ## Installation 1. Clone the repository: ```bash git clone github.com cd naija-tax-agent ``` 2. Install dependencies using uv: ```bash make install ``` This will automatically: - Install uv if not already installed - Sync Python dependencies using uv ## Development ### Running the Project There are several make commands available for development: ```bash # Start the API server make dev # Launch the playground interface make playground ``` The API server runs with CORS enabled for all origins, while the playground interface is available on port 8501. ## Project Structure ``` naija-tax-agent/ ├── Makefile # Build automationdependencies ├── README.md # Project documentation ├── app/ # Main package directory | ├── pyproject.toml # Project metadata and | ├── naija-tax/ # Main package directory | │ ├── __init__.py │ │ ├── agent.py # Main agent implementation │ │ ├── prompt.py # Agent prompts and instructions │ │ ├── tools/ # Agent tools directory │ │ │ ├── cac_public_search.py │ │ │ └── search.py ├── frontend/ # Frontend application | ├── README.md …