querybuilder-llm

A simple pip package that lets you generate SQL or MongoDB queries from natural language using your deployed LLM API.
Schema is passed in, and the LLM figures out the correct query.
All responses are wrapped with a success flag for safety.

Installation

pip install querybuilder-llm

Usage

from querybuilder_llm.client import build_query

# Example schema
schema = {
    "users": ["id", "name", "email", "age"]
}

result = build_query(schema, "Find all gmail users", "postgresql")

if result["success"]:
    print("SQL Query:", result["response"])
else:
    print("Error:", result["error"])

# --- Mongo Example ---
result = build_query(schema, "Find users older than 30", "mongodb")

if result["success"]:
    print("Mongo Query:", result["response"])
else:
    print("Error:", result["error"])

Response Format

{
  "success": True/False,
  "response": str | dict | None,
  "db_type": postgress|mongodb
  "error": str | None
}

success: Whether query generation was successful

response: SQL string or MongoDB JSON dict (if success)

error: Error details if failed