# Simple Hierarchical AI Shopping‑List Agent

## What it does
This repository contains a minimal example of a **hierarchical LangChain agent** that helps you plan a shopping list. The main agent:

1. Accepts a natural‑language request with a list of products and a city.
2. Calls an internal tool `get_price(product, city)` which itself creates a tiny sub‑agent to generate realistic price tables for the requested product in the specified city.
3. Aggregates the prices and prints a final summary.

The example demonstrates:
- Connecting to a local LLM via LM Studio (OpenAI‑compatible API).
- Using `@tool` to expose a function that internally runs another agent.
- Running the main agent with `create_agent` and printing all intermediate tool calls.

## File structure
| File | Purpose |
|------|---------|
| `requirements.txt` | Pinning LangChain, langchain‑openai and dotenv dependencies |
| `main.py` | Full implementation of the hierarchical agent and demo usage |
| `README.md` | This documentation |

## Installation
```bash
python -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install -r requirements.txt
```
Make sure LM Studio is running on `http://localhost:1234/v1` (or set `LM_BASE_URL` and `LM_MODEL`).

## Running the demo
```bash
python main.py
```
The script will:
- Run the main agent with a hard‑coded prompt.
- Print each tool call (`get_price`) and its output.
- Show aggregated price tables and total cost.

Feel free to modify `user_prompt` in `main.py` or pass input from stdin for interactive use.