Booknauu API Test Framework

AI-Powered Automated API Testing for the Booknauu Platform

A test automation framework built from scratch for the Booknauu backend API. Designed with a minimalist approach, it leverages a local AI model (Ollama) to generate dynamic, realistic test data and includes both API and database integration verification.

---

What This Framework Achieved (Measurable Results)

• 70+ automated test scenarios covering 9 critical business modules (Authentication, Resource Management, Reservations, Team, Dashboard, Google Integration, and more).
• 8 critical bugs identified during the first test run, including:
  • 500 Internal Server Errors on reservation and resource endpoints.
  • 404 misconfigurations on protected endpoints.
  • Inconsistencies between OpenAPI documentation and actual backend responses.
• 95% test setup reliability achieved with a built‑in retry mechanism that handles 409 Conflict errors by generating unique data on the fly.
• Reduced manual test data preparation time by ~80% using AI‑generated values (email, business names, resource names, etc.) combined with timestamp‑based uniqueness.
• Extended the framework with a Database Integration layer (PostgreSQL + SQLAlchemy) to directly verify data persistence, cutting debugging time by ~50%.

---

Key Features

• AI-Powered Data Generation – Uses Ollama (LLaMA 3.2) to automatically generate values such as email, businessName, resourceName, and more.

• Automatic Test Context Setup – A global fixture performs Register → Login → Create Resource Type → Create Resource → Create Reservation before the tests run, so every test starts with a clean and consistent environment.

• Comprehensive API Test Coverage – Covers 70+ scenarios across 9 modules, including Auth, Resources, Reservations, Google Integration, Team, Dashboard, and Operating Hours.

• Conflict Retry Mechanism – If a 409 Conflict occurs (duplicate email), the framework automatically retries with a newly generated email, improving test stability.

• HTML Reports – Generates visual test reports with pytest-html for easier analysis.

• Multi‑Layer Testing:

  • API Layer – Validates HTTP responses and status codes.
  • Integration Layer – Checks that data is correctly persisted in the PostgreSQL database.
  • E2E Layer – (Planned) UI tests with Playwright.

---

Tech Stack

Component	Technology
Language	Python 3.11+
Test Runner	Pytest
HTTP Client	Requests
AI Inference	Ollama (LLaMA 3.2 3B)
Database Integration	SQLAlchemy + psycopg2‑binary
Reporting	pytest‑html
Environment	python‑dotenv

---

Prerequisites

Before running the framework:

1. Booknauu Backend must be running at http://localhost:8080.

2. Ollama must be installed and running with the llama3.2:3b model.

   ollama pull llama3.2:3b
   

Notes

• Google Calendar tests are skipped by default because they require OAuth credentials (Client ID, Secret) which are not set in the local environment.

• Integration tests require a running PostgreSQL database. They directly verify that data is correctly saved after API calls.

• All numbers (70+ scenarios, 8 bugs, 95% reliability, etc.) are based on actual test execution results.