Basic AI Agent

A local AI agent project powered by Ollama, LangChain, and LangGraph. The agent runs on your machine, can use tools, remembers conversation context, and includes both command-line and web interfaces.

Features

• Local LLM support through Ollama
• ReAct-style agent workflow using LangChain and LangGraph
• Calculator tool for math expressions
• Knowledge base tool for technical topics
• Date and time tool
• Conversation memory support
• Streamlit chat UI
• FastAPI backend for API-based chat
• React + Vite frontend
• Unit tests for tools

Project Structure

Basic_Agent/
+-- agent.py                  # Basic agent example
+-- agent_with_memory.py      # Agent with conversation memory
+-- agent_langgraph.py        # LangGraph-based agent
+-- app.py                    # Streamlit web app
+-- config.py                 # Environment-based configuration
+-- setup.py                  # Local setup checker
+-- requirements.txt          # Python dependencies
+-- backend/
|   +-- api.py                # FastAPI backend
+-- frontend/
|   +-- package.json          # React frontend dependencies
|   +-- src/                  # Frontend source files
+-- tools/
|   +-- calculator.py
|   +-- datetime_tool.py
|   +-- knowledge_base.py
+-- tests/
    +-- test_tools.py

Requirements

• Python 3.9 or newer
• Node.js and npm, for the React frontend
• Ollama installed locally
• An Ollama model such as phi3, llama3.2, or mistral

Setup

Clone the project and enter the folder:

git clone https://github.com/YOUR_USERNAME/Basic_Agent.git
cd Basic_Agent

Create and activate a virtual environment:

python -m venv venv
venv\Scripts\activate

Install Python dependencies:

pip install -r requirements.txt

Install Ollama from:

https://ollama.com/download

Pull a local model:

ollama pull phi3

Start Ollama:

ollama serve

Configuration

Create a .env file in the project root:

OLLAMA_MODEL=phi3
OLLAMA_BASE_URL=http://localhost:11434
TEMPERATURE=0.1
MAX_TOKENS=1024

The .env file is ignored by Git, so your local settings stay private.

Run The Agent

Run the basic agent:

python agent.py

Run the memory-enabled agent:

python agent_with_memory.py

Run the LangGraph agent:

python agent_langgraph.py

Run the Streamlit web UI:

streamlit run app.py

Run The FastAPI Backend

If needed, install API dependencies:

pip install fastapi uvicorn pydantic

Start the API server:

uvicorn backend.api:app --reload --port 8000

Useful API endpoints:

GET  /health
GET  /tools
POST /chat
GET  /sessions
GET  /sessions/{session_id}
DELETE /sessions/{session_id}

Run The React Frontend

Open a second terminal:

cd frontend
npm install
npm run dev

The frontend usually runs at:

http://localhost:5173

Make sure the FastAPI backend is running on port 8000 before using the frontend chat.

Tools

Tool	Purpose	Example
Calculator	Evaluates math expressions	sqrt(144), 2**10
Knowledge Base	Looks up technical topics	python, langchain, ollama
DateTime	Returns current date and time	date, time, full

Run Tests

python -m pytest tests -v

GitHub Push Commands

If this is a new repository:

git init
git branch -M main
git add .
git commit -m "Initial commit"
git remote add origin https://github.com/YOUR_USERNAME/Basic_Agent.git
git push -u origin main

Replace YOUR_USERNAME with your GitHub username.

Notes

• Do not commit .env, venv/, __pycache__/, frontend/node_modules/, or build output.
• Keep Ollama running before starting the agent.
• Use a model that works well with tool calling for best results.