QA-Kit

QA-Kit is a CLI and test generator for REST API integration tests using pytest and httpx, with built-in support for Allure reporting. Generate tests from JSON specifications, run them asynchronously, and generate detailed HTML reports.

---

Documentation

For detailed usage instructions, CI/CD examples, and setup guide, see the full documentation:

QA-Kit User Guide (Google Docs)

Allure Report

https://roshanguptamca.github.io/qa-kit/

Features

• Generate pytest tests from JSON API specifications
• Async HTTP requests with httpx
• Partial JSON assertions with recursive key ignore support
• Optional wildcard matching for ignored keys
• Global option to skip assertions
• Allure reporting support (HTML reports)
• Fully configurable via environment variables
• CLI for generation, running, linting, and reporting
• Production-ready and PyPI compatible

---

Installation

1. Clone the repository

git clone https://github.com/roshanguptamca/qa-kit.git
cd qa-kit

Install Python dependencies (recommended via Poetry)

poetry install

Install Allure CLI (required for HTML reports)

macOS (Homebrew):

brew install allure

Linux (Debian/Ubuntu):

sudo apt-add-repository ppa:qameta/allure
sudo apt update
sudo apt install allure

Windows (via Scoop):

scoop install allure

Verify installation:

allure --version

⚠️ QA-Kit will fail to generate HTML reports without the Allure CLI.

4. (Optional) Docker

If you prefer, Allure reports can also be generated via Docker:

docker run --rm -v $(pwd)/allure-results:/allure-results -v $(pwd)/allure-report:/allure

Quick Start

pip install qa-kit


Or install from GitHub:

```bash
pip install git+https://github.com/roshanguptamca/qa-kit.git

1. Prepare JSON specs Example: tests/specs/cart_api.json: json { "name": "Cart API Suite", "base_url": "https://api.example.com", "tests": [ { "id": "create-cart-1", "name": "create_cart", "method": "POST", "path": "/cart/", "body": { "channel": {"id": "online", "name": "Online"} }, "expected": { "status_code": 200, "json": { "status": "ACTIVE", "channel": {"id": "online"} } } } ] }

    ```
   

2. Generate tests

   qa_kit generate tests/specs/cart_api.json -o tests/generated
   

3. Run tests with Allure reporting ```bash` qa_kit run -t tests/generated

4. Generate Allure report

   qa_kit report -o allure-report
   

This runs tests and generates results in allure-results. You can then open the report: bash qa_kit open

Environment Variables

• QA_KIT_SSL_VERIFY: Set to true to enable SSL verification (default: false)
• USE_WILDCARD: Set to true to enable wildcard matching for ignored keys (default: false)
• IGNORE_ASSERT : Set to true to skip JSON assertions (default: false)

JSON Spec Options:

Field	Description
id	Unique test identifier
name	Test name
method	HTTP method (GET, POST, etc.)
path	Endpoint path
body	JSON body for request
params	Query parameters
expected.status_code	Expected HTTP status code
expected.json	Expected JSON response
ignore_assert	Skip JSON assertions for this test
ignore_json	List of keys to ignore recursively in assertions
use_wildcard	Enable wildcard for ignored keys

Example Generated Test:

import pytest
import allure
import httpx
from tests.utils.test_helpers import _assert_partial, SSL_VERIFY

BASE_URL = "https://jsonplaceholder.typicode.com"

@allure.story("JSONPlaceholder API Suite")
async def test_get_post_1_get_post_1():
    """get_post_1"""
    async with httpx.AsyncClient(base_url=BASE_URL, verify=SSL_VERIFY) as client:
        resp = await client.request(
            "GET",
            "/posts/1",
            json={},
            params={}
        )

    assert resp.status_code == 200
    _assert_partial({
        "userId": 1,
        "id": 1
    }, resp.json(), ignore_keys=[], use_wildcard=False)

Example:

Project Structure

tests/
├── specs/         # JSON API specs
├── generated/     # Generated tests
└── utils/
    └── test_helpers.py  # Shared helpers

Contributing

Contributions welcome! Please fork, create a branch, and submit PRs. Follow black, isort, and flake8 for code formatting and linting.

QA-Kit Roadmap:

OpenAPI / Swagger Spec Support

 Input: JSON/YAML OpenAPI spec
 Output: Auto-generated async pytest tests
 Optional: Generate default request bodies based on schema

Parametrized & Data-Driven Tests

Support for multiple test cases from JSON, CSV, or Excel
Generate @pytest.mark.parametrize automatically

AI-Assisted Test Generation

Focus: Help users automatically generate tests based on API specs or historical data.

Features: Smart test generation:

    Suggest additional test cases (edge cases, missing inputs) based on spec analysis.
    
    Automatically generate negative test scenarios (invalid payloads, bad params).
    
    AI-based validation hints:
    
    Suggest keys to ignore or fields to focus on in assertions.
    
    Integration with OpenAI / local LLM:
    
    Generate descriptive docstrings for tests.
    
    Provide inline suggestions for improving test coverage.

Example CLI:
    qa_kit --ai-assist --spec api_spec.json --output tests/generated
    qa_kit --ai-assist--test tests/generated/test_api.py

AI-Powered Test Optimization

Features:

Test suite prioritization:
    Rank tests based on likelihood of failure, coverage gaps, or historical flakiness.
    Duplicate test detection:
    Automatically detect overlapping or redundant tests and suggest merges.
    Response anomaly detection:
    Use AI to detect unusual response patterns in API results, highlighting potential bugs.

AI-Driven Reporting & Analysis

Features:

Smart report summaries:
    Auto-generate concise human-readable summaries for QA/management.
    Trend analysis:
        Predict potential failure trends across builds or releases using historical data.
        Code & spec improvement suggestions:
        Suggest changes to API specs or test payloads based on AI analysis of failed tests.
    Example Output:
        Highlight which endpoints are high-risk
        Predict areas with insufficient test coverage
        Suggest new tests for endpoints with historical instability

Fully Autonomous Test Assistant

Features:

Continuous learning:
    Autonomous test suite maintenance:
        Automatically updates or generates new tests for spec changes.
    Auto-fix broken tests:
        Proposes corrections for failing tests due to schema changes or new endpoints.
    Natural language interface:
        Users can describe tests in plain English; AI converts to pytest tests.
    Example CLI:
       qa_kit ai --describe "Test user creation with invalid email" --generate

CLI Enhancements

qa_kit validate-spec → check spec file validity
 qa_kit diff-tests → compare generated tests with previous versions
 Improved qa_kit clean → safely clean tests and reports

Improved Report Handling

Auto-generate Allure reports if results exist
Option to publish HTML to GitHub Pages
Add trend charts and summary tables in the report

Better Async Framework Support

 Support for aiohttp + optional sync mode with requests
 Users can choose framework via CLI flags

CI/CD Integration

Predefined GitHub Actions, GitLab CI, and Jenkins pipelines
Docker-ready templates for isolated test runs
Postman / HAR Import
Import Postman collections / HAR files as JSON specs

Generate tests automatically

Mock Server / Sandbox Mode
Optional mock server for testing without live APIs
Record & replay responses

Soft Assertions

Allow multiple assertion failures in one test run
Configurable via CLI flags

Advanced Reporting & Extensibility

Plugin System
    Users can write plugins for custom generators, validators, or reporters
    Simple interface to extend functionality
Enhanced Reports
    PDF export of Allure results
    Include request/response snapshots
    Visual trend charts for API performance
Pre/Post Test Hooks
    Users can define setup/teardown hooks for each test or suite
    Ideal for authentication, DB reset, or test data setup

Production-Ready Ecosystem

Packaging & Distribution
    Publish to PyPI for easy installation
    Docker images for isolated environments
Comprehensive Documentation
    User guide, API reference, and examples
    Tutorials for common use cases
Community & Support
    Dedicated forum or Discord for user support
    Regular updates and maintenance

License

MIT © RoshanGupta