🧠 LangGraph SQL Agent

This project demonstrates how to build a powerful SQL Agent using LangGraph and LangChain. It enables an LLM (like OpenAI's GPT models) to interact with a SQL database through multi-step reasoning, tool use, and error correction

---

🚀 Features

• 🔍 Natural Language to SQL: Ask questions in plain English and get accurate SQL queries generated and executed
• 📊 Schema-Aware: Automatically inspects the database schema before querying
• 🧠 Multi-step Agentic Reasoning: Uses LangGraph to allow agents to iterate, check, and fix themselves
• 🔁 Error Correction: When queries fail, the agent reflects, debugs, and retries

---

🧰 Tech Stack

Tool	Purpose
LangGraph	Agent Workflow Orchestration Framework
LangChain	LLM framework with SQL support tools
SQLite	Lightweight ORM database
Python	Core programming language
OpenAI API	Language model backend

---

🧭 Agent Workflow Explained

The above diagram illustrates the LangGraph SQL Agent's reasoning process — a step-by-step flow enabling natural language to SQL execution through LLM-powered tools:

🔄 Step-by-Step Breakdown

• __start__
  Entry point of the graph-based agent.

• list_tables
  Retrieves a list of all tables in the database to give the agent context about available data.

• call_get_schema
  Triggers the retrieval of detailed schema information for specific tables.

• get_schema
  Gathers column names, data types, and structure — critical for accurate SQL generation.

• generate_query
  The LLM generates a SQL query based on the user's natural language input and the retrieved schema.

• check_query (optional validation step)
  Evaluates the SQL query. If an error is detected, it loops back to refine or regenerate the query.

• run_query
  Executes the generated query on the connected database and collects the result.

• __end__
  Final state of the graph. Either returns a successful result or exits gracefully if an issue occurs.

---

📁 File Structure

├── sql-agent.ipynb        
├── requirements.txt       
└── README.md              

---

📦 Setup Instructions

1. Clone the Repository

   git clone https://github.com/your-username/sql-agent.git
   cd sql-agent

2. Install Dependencies You can use pip or a virtual environment:

   pip install -r requirements.txt

3. Configure Environment Variables Set your OpenAI API Key:

      OPENAI_API_KEY=your-key-here    

---

📝 Example Prompts

• "What is the average salary in the employees table?"
• "List all customers who made a purchase in the last 30 days"
• "How many orders were placed in each region?"

---

⚠️ Limitations

• Currently supports read-only SQL operations (i.e., SELECT)
• For INSERT, UPDATE, or DELETE, you'd need to build custom tools

---

🧩 Extending the Project

• Build custom LangChain tools for SQL INSERT, UPDATE, etc.
• Integrate schema validation for safer write operations
• Add UI with Streamlit

---