Coding AI agent using Pydantic AI's Direct API for minimal abstraction. Features 9 file system tools and interactive terminal interface.

Codantic AI - Educational AI Agent Loop

For Educational Purposes Only - Learn AI agent architecture with minimal abstractions

A Python AI agent system using Pydantic AI's Direct API for maximum learning clarity. Built with Google Gemini and featuring 9 file system tools.

🧠 Why This Approach?

Uses pydantic_ai.direct.model_request_sync() - minimal abstraction perfect for learning:

• 🔍 See exactly how LLM interactions work
• 📚 No magic - every API call is explicit
• 🛠 Direct control over context and tool calling

🎬 Live Demo

🛠 9 Built-in Tools

🚀 Quick Start

Prerequisites

• Python 3.13+
• Google Claude API key (Get one here)

Installation

git clone <your-repo-url>
cd codantic-ai

# Setup environment
python -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate

# Install dependencies  
pip install -r requirements.txt
# OR: uv sync

# Configure API key
cp .env.example .env
# Edit .env and add your Claude API key

# Run the agent
python main.py

🏗 How It Works

graph TB
    A[👤 User Input] --> B[🔄 AgentLoop]
    B --> C[🤖 Pydantic AI Direct API]
    C --> D{🛠 Tool Calls?}
    D -->|Yes| E[⚡ Execute Tools]
    E --> B
    D -->|No| F[💬 Response]

Key Learning Points

# Direct API usage - no hidden abstractions!
model_response = model_request_sync(
    self.model_name,
    self.context,  # Full conversation history
    model_request_parameters=ModelRequestParameters(
        fun