Simple LLM Agent (with tools usage)

This is a library which contains boilerplate code/basic example of different ways to create your AI agent that uses custom tools and locally run llm model (ollama). You can easily modify it to use any other model as well.

Examples:

1. Langgraph Agent - Uses Langgraph library and related langchain sum library for a basic reactive agent with simple tools.
2. Strands Agent - Uses Strands-Agent library and strands-agents-tools
3. More to come.

Prerequisite

uv

It can be installed using pip install uv

use command uv --help to check if it is installed correctly.

Ollama

Install ollama - Install ollama model with tooling support (https://ollama.com/blog/tool-support, https://ollama.com/search?c=tools)

• ollama run qwen3:4b pr ollama run llama3.2 It opens chat terminal which you can test for chatting or just close using /bye.

Start ollama server

• ollama serve

Installing libraries

uv add langgraph langchain_core langchain langchain-ollama

Running the agent

Comment/Uncomment the code to run specific Agent example in main.py. Then run the main module main.py using uv run -m main

or

.\.venv\Scripts\python.exe main.py

or to just run the langgraph module uv run -m langgraph_chat.chat_assitant