Prompt Chain

Validate SQL Queries Against Expected Outputs

Name: Validate SQL Queries Against Expected Outputs
Availability: OnlineOnly
Author: Promptfoo

A multi-step prompt workflow that validates SQL queries for correctness, security vulnerabilities, and best practices before execution in production

Copy chain

Works with github

Promptfoo

Maintainer?

Spark score

out of 100

Updated yesterday

Version code-scan-action-0.1

Add to Favorites

Why it matters

Ensure the accuracy and correctness of your SQL queries by automatically validating their output against predefined expectations. This asset helps catch errors and regressions before they impact your data.

Outcomes

What it gets done

Execute SQL queries against a database.

Compare query results to expected outcomes.

Identify discrepancies and potential errors in SQL.

Automate the testing of SQL query logic.

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/pfoo-sql-validation | bash

Capabilities

What this chain does

Query a database

Writes and executes SQL or NoSQL queries on databases.

Write tests

Creates unit, integration, or end-to-end test cases.

Debug

Traces errors to their root cause and suggests fixes.

Extract

Pulls structured data fields from unstructured text.

Overview

Sql Validation

What it does

This prompt chain implements a structured, multi-step workflow for validating SQL queries. It systematically evaluates queries for syntax correctness, security vulnerabilities like SQL injection, and adherence to performance best practices. The workflow is designed to run as an automated validation pipeline using the promptfoo framework.

How it connects

Use this when you need to validate SQL queries before production deployment, especially in environments where untrusted input may influence query construction. It's ideal for teams building database-backed applications who want to catch SQL errors, security risks, and performance issues during development rather than in production.

Source README

yaml-language-server: $schema=https://promptfoo.dev/config-schema.json

description: 'Is-SQL example'

prompts:

Output a SQL query that returns the number of {{product}} sold in the last month.

Database schema:
{{database}}

Only output raw SQL code without markdown.

providers:

openai:gpt-4.1-mini
openai:gpt-4o

tests:

vars:
product: bananas
database: file://database.sql

assert:

Ensure output is a valid SQL query
- type: is-sql
- type: icontains
  value: 'Select'
vars:
product: apples
database: file://database.sql
assert:

Ensure output is a valid SQL query
- type: is-sql
  value:
  databaseType: 'MySQL'
  Ensure only the following tables are used
  allowedTables:
  - select::null::Products
  - select::null::Shipments
vars:
product: oranges
database: file://database.sql
assert:
- type: is-sql
  value:
  databaseType: 'MySQL'
  allowedColumns:
  # We expect this test to fail no matter what
  - select::null::DoesntExist

Discussion