k11TechLab Vibium Jest AI Test Framework

A production-style test automation framework for https://k11softwaresolutions.com featuring an AI agent and MCP server integration, built with Jest, Vibium, structured reporting, Lighthouse auditing, AI-assisted test generation, and GitHub Actions CI/CD.

Highlights

• Shared Vibium browser setup with reusable page objects
• Coverage across smoke, functional, API, DB, devices, E2E, and performance validation
• Timestamped HTML and JSON reports in reports/
• AI-assisted test generation with optional OpenAI integration
• AI agent and MCP server integration for tool-driven browser automation
• GitHub Actions CI for API and smoke validation with artifact upload
• GitHub Actions release workflow for showcase bundle packaging
• MIT licensed and production-ready

Capabilities & Innovations

• Vibium + TypeScript + Jest: Tests, shared setup, page objects, and helpers are implemented in TypeScript. Page objects live under src/k11-platform/pageObjects/, shared browser lifecycle is handled in src/k11-platform/hooks/, and reusable helpers live in src/utils/.
• Mixed-layer automation: The framework covers smoke, functional, API, DB, device, E2E, Lighthouse, and AI-generated test flows in a single repository.
• Lighthouse automation: Lighthouse runs generate HTML and JSON outputs in artifacts/lighthouse/ for performance-focused validation and demo-ready reporting.
• DB & API helpers: Reusable utilities in src/utils/ support API validation, DB access, and test data handling.
• AI-assisted test generation: A repo-aware generator in scripts/ai/ can inspect page objects, build Jest-compatible Vibium tests, and optionally run them immediately.
• Prompt and runner ecosystem: Root-level prompts/ contains reusable generation goals, and ai-test-runner/ provides one-click shell, PowerShell, CMD, and npm runner options.
• Agent + MCP demo: mcp-server/ and agent/ demonstrate how an AI agent can call Vibium-backed browser tools through a real stdio MCP server.
• Advanced reporting: Timestamped HTML and JSON Jest reports are written to reports/, with screenshots and Lighthouse artifacts captured under artifacts/.
• CI / CD: GitHub Actions workflows in .github/workflows/ run validation suites, upload artifacts, and package a release-ready showcase bundle.
• Clean demo output: Opt-in debug logging via K11_DEBUG=true keeps normal runs polished while preserving diagnostics when needed.
• OpenAI integration with fallback mode: The AI generator uses the OpenAI Responses API when configured and falls back to a local repo-aware template when AI access is unavailable.

Framework Architecture

src/k11-platform/
|-- config/        app URLs, browser, timeouts
|-- hooks/         shared Vibium lifecycle setup
|-- pageObjects/   reusable page models and actions
|-- testdata/      CSV and other input data
|-- tests/
|   |-- smoke/
|   |-- functional/
|   |-- api/
|   |-- db/
|   |-- devices/
|   |-- e2e/
|   |-- lighthouse/
|   |-- generated/
|-- apiResponses/  response payload helpers
src/utils/         data, DB, API, UI, and debug helper utilities
scripts/
|-- ai/            AI test generator and prompt-driven creation flow
|-- vibium-demo/   lightweight Vibium demos for humans and agents
|-- jest-root-report.cjs  Jest HTML/JSON report writer
agent/             OpenAI-driven MCP client runner
mcp-server/        Vibium-backed MCP tool server
ai-test-runner/    one-click AI generation runners
prompts/           reusable goal prompts for AI generation
artifacts/         screenshots, Lighthouse outputs, and demo assets
reports/
|-- *.html/*.json  timestamped Jest reports
|-- ai-agent/      timestamped AI agent HTML/JSON run reports
.github/workflows/ CI and release workflows

Execution Flow

1. Jest starts a suite
2. setupVibium() launches the browser and page
3. Tests call shared page objects from src/k11-platform/hooks/vibiumSetup.ts
4. Assertions run on live UI, API, DB, or performance flows
5. teardownVibium() closes the browser
6. Reports and artifacts are written to disk

Tech Stack

• Jest for test execution and mixed-layer orchestration
• Vibium for browser automation
• TypeScript for framework, page objects, helpers, and generated tests
• OpenAI Responses API for AI-assisted test generation and the demo agent
• Model Context Protocol (MCP) over stdio for tool-driven agent integration
• Lighthouse for performance audits
• GitHub Actions for CI/CD

Prerequisites

• Node.js 18+
• Python 3.9+
• Vibium runtime and browser dependencies

Install

JavaScript

npm install

Python

python -m pip install -r requirements.txt

Run

Run all tests:

npm test

Run targeted suites:

npm run test:smoke
npm run test:functional
npm run test:api
npm run test:db
npm run test:devices
npm run test:e2e
npm run test:lighthouse

Run the CI suite locally:

npm run test:ci

Run legacy Vibium demos:

npm run vibium:k11:async
npm run vibium:k11:sync
python scripts/vibium/k11_smoke.py

Agentic MCP Demo

Run the Vibium MCP server:

npm run mcp:server

Run the demo agent:

npm run agent:run -- "Open the K11 homepage and verify the contact page is reachable."

See mcp-server/README.md and scripts/vibium-demo/README.md for the MCP and agentic demo details.

AI-Assisted Generation

Create a starter Jest + Vibium test:

npm run ai:generate:test -- --page-object LoginPage --goal "Verify valid login reaches dashboard"

Generate and run immediately:

npm run ai:generate:test -- --page-object HomePage --goal "Verify the hero and navigate to services" --run

Use preset one-click runners:

npm run ai:runner:homepage
npm run ai:runner:login
npm run ai:runner:formslab

The generator auto-loads .env from the repo root if present. Start from .env.example for OpenAI configuration.

See doc/ai-test-generator.md for generator details.

CI/CD

GitHub Actions workflows are included in .github/workflows/:

• ci.yml Runs on push, pull request, and manual trigger. Installs Node and Python dependencies, runs API and smoke suites, and uploads reports and artifacts.

• release-showcase.yml Runs on version tags like v1.0.0 or via manual trigger. Packages the showcase framework into a zip file and publishes it as a workflow artifact. On tag pushes, it also creates a GitHub release.

Reports And Artifacts

Generated outputs are written to:

• reports/ for timestamped Jest HTML and JSON reports
• artifacts/screenshots/ for UI screenshots
• artifacts/lighthouse/ for Lighthouse HTML and JSON outputs

Documentation

Additional docs are available in doc/, including:

• doc/ai-assisted-test-automation.md
• doc/ai-test-generator.md
• doc/ai-test-generator-functions.md
• doc/jest-vs-fixtures.md
• doc/jest-vs-playwright-test.md
• doc/reporting-implementation.md
• doc/mcp-stdio.md
• doc/vibium-features.md

Additional runnable demos and agent docs:

• scripts/vibium-demo/README.md
• mcp-server/README.md

Acknowledgements

This framework was also refined with Codex, used as an AI engineering assistant for implementation, debugging, documentation, and framework acceleration.

License

This project is licensed under the MIT License. See LICENSE.

Author

Kavita Jadhav

Accomplished Full Stack Developer and Test Automation Engineer specializing in modern web application development, robust full stack solutions, and scalable automation frameworks. Expert in Playwright, advanced quality engineering, and driving best practices for high-impact, reliable software delivery.

LinkedIn: https://www.linkedin.com/in/kavita-jadhav-tech/

About k11 Software Solutions

k11 Software Solutions is a leading provider of modern, AI-powered test automation, DevOps, and quality engineering services. We help organizations accelerate digital transformation with robust, scalable, and intelligent automation solutions tailored for SaaS, web, and enterprise platforms.

• Website: https://k11softwaresolutions.com
• Contact: k11softwaresolutions@gmail.com

Partner with us to future-proof your QA and automation strategy!

Follow Me