ThinAgent (LangGraph) — Calculator Agent

A beginner-friendly thin agent built with LangGraph + LangChain.
It reads a natural-language math request (e.g., "multiply 6 by 7 then add 4"), decides which tool to call (add/sub/mul/div), and can optionally reflect (review its own answer) up to 2 times for quality.

---

✨ What this Calculator Agent does

• Parses plain-English math prompts.
• Uses an LLM to decide which tool(s) to call:
  • add(num1, num2), sub(...), mul(...), div(...)
• Orchestrates the flow with LangGraph:
  • Nodes: LLM → Tools → (optional) Reflection → Final Answer
  • Conditional edges choose whether to run tools again, reflect, or finish.

---

🧱 Suggested project structure

ThinAgent_Langgraph/
├─ calculator_agent.py      # Main agent (LangGraph) + tools + demo run
├─ requirements.txt         # Python deps
├─ test_project.py          # Integration–style tests for the agent
├─ .env.example             # Template for env vars (no secrets)
├─ .gitignore               # Ignore .env, caches, etc.
└─ README.md

---

✅ Prerequisites

• Python 3.10+
• An OpenAI API key (or compatible provider)
• Git

---

🚀 Setup (step-by-step)

1) Clone the repo

git clone https://github.com/selvicim45/ThinAgent_Langgraph.git
cd ThinAgent_Langgraph

2) Create & activate a virtual environment

macOS / Linux

python3 -m venv .venv
source .venv/bin/activate

Windows (PowerShell)

python -m venv .venv
.venv\Scripts\Activate.ps1

3) Install dependencies

Create requirements.txt (if not present) with:

python-dotenv
langchain
langchain-openai
langgraph
openai
pytest

Install:

pip install --upgrade pip
pip install -r requirements.txt

Testing

Run all tests:

pytest -q