{"name":"AI-Agent-Test-Using-Amazon-Nova-Act","generatedAt":"2026-10-04T08:25:35.712Z","languages":{"Python":45420},"commitTotal":4,"commits":[{"sha":"6789d47d26d9cde9e11c1b0a30e29805f4bb3498","message":"Remove Lottie.js script tag from README\n\nRemove unused script tag for Lottie animation.","date":"2026-03-07T12:01:31Z","author":"Aayush Mishra","login":"Aayush-Mishraa","avatar":"https://avatars.githubusercontent.com/u/89012492?v=4","verified":true,"url":"https://github.com/Aayush-Mishraa/AI-Agent-Test-Using-Amazon-Nova-Act/commit/6789d47d26d9cde9e11c1b0a30e29805f4bb3498"},{"sha":"94e56ae4f9abbc862015773ae075bfc0a8003718","message":"Update README.md","date":"2026-03-07T11:56:41Z","author":"Aayush Mishra","login":"Aayush-Mishraa","avatar":"https://avatars.githubusercontent.com/u/89012492?v=4","verified":true,"url":"https://github.com/Aayush-Mishraa/AI-Agent-Test-Using-Amazon-Nova-Act/commit/94e56ae4f9abbc862015773ae075bfc0a8003718"},{"sha":"34512dba010a3bb694e27814c2c20f7e701be4d5","message":"Add professional project README","date":"2026-03-07T09:00:44Z","author":"Aayush Mishra","login":"Aayush-Mishraa","avatar":"https://avatars.githubusercontent.com/u/89012492?v=4","verified":false,"url":"https://github.com/Aayush-Mishraa/AI-Agent-Test-Using-Amazon-Nova-Act/commit/34512dba010a3bb694e27814c2c20f7e701be4d5"},{"sha":"25de46a5f405b9b624392266ec697eae61179f87","message":"Add Nova Act scenarios and HTML test reporting","date":"2026-03-07T08:48:57Z","author":"Aayush Mishra","login":"Aayush-Mishraa","avatar":"https://avatars.githubusercontent.com/u/89012492?v=4","verified":false,"url":"https://github.com/Aayush-Mishraa/AI-Agent-Test-Using-Amazon-Nova-Act/commit/25de46a5f405b9b624392266ec697eae61179f87"}],"readmeHtml":"

AI-Test-Automation-Using-Amazon-Nova-Act

\n

1. Introduction

\n

This project shows how to build AI-driven web test automation using Amazon Nova Act. Instead of writing rigid locator-heavy scripts for every UI action, we describe test intent in natural language and let Nova Act execute those steps in the browser.

\n

The goal of this repository is practical: demonstrate real scenario-based functional testing for an ecommerce flow using AI agents, while still keeping the project structured like an engineering test suite.

\n

2. Problem with Traditional Automation

\n

Traditional UI automation frameworks like Selenium, Playwright, and Cypress are powerful, but they come with known pain points in large test suites:

\n\n

In short, traditional tools are deterministic and robust, but often expensive to maintain over time for fast-moving UI products.

\n

3. Why AI Driven Automation

\n

AI-driven automation shifts from selector-first scripts to intent-first instructions.

\n

Instead of writing low-level code for every click and input, we provide high-level instructions such as:

\n

\"Login with valid credentials and verify username on dashboard\"

\n

An AI agent can then:

\n\n

This approach can reduce authoring effort and improve adaptability for exploratory and regression-style flows.

\n

4. Amazon Nova Act Overview

\n

Amazon Nova Act is an AI agent capability from AWS for browser task automation. It combines LLM reasoning with an action execution layer to perform tasks on real web pages.

\n

At a high level, Nova Act:

\n\n

In this project, Nova Act is used through the Python SDK with workflow-based scenario execution.

\n

5. Architecture

\n

The implementation follows an agentic test architecture where test intent is translated into executable browser actions.

\n

Core building blocks:

\n\n

Architecture flow:

\n
User Prompt\n  |\n  v\nNova LLM\n  |\n  v\nAgent Planner\n  |\n  v\nBrowser Automation\n  |\n  v\nExecution Logs\n
\n

6. How Nova Act Works Internally

\n

In simple terms, each nova.act(...) instruction goes through this cycle:

\n
    \n
  1. Parse intent from natural language.
    2. \n
  2. Observe current browser/page state.
    3. \n
  3. Decide the next best action.
    4. \n
  4. Execute the action.
    5. \n
  5. Re-evaluate whether the goal is complete.
    6. \n
  6. Continue until the instruction is satisfied or fails.
    7. \n
\n

This loop is why AI automation can handle small UI differences better than hardcoded scripts.

\n

7. Browser Interaction

\n

Nova Act uses a browser automation layer compatible with Playwright-style interaction patterns under the hood. This enables stable low-level control (click, type, navigate, wait, evaluate) while exposing a high-level natural-language interface to the test author.

\n

In practice, you write intent, and the framework handles granular browser operations.

\n

8. Nova Act Playground

\n

The Nova Act Playground is a browser-based environment for quickly trying prompts and observing agent behavior without writing much code.

\n

What it gives you:

\n\n

Regional availability note:

\n\n

If playground access is limited, use the SDK directly from local scripts, which is the approach used in this repository.

\n

9. Nova Act SDK Setup

\n

Follow these steps to run locally.

\n

9.1 Prerequisites

\n\n

9.2 Install Dependencies

\n
pip install -r requirements.txt\npip install playwright\nplaywright install
\n

9.3 Install Nova Act SDK Explicitly (if needed)

\n
pip install nova-act
\n

9.4 Create IAM User for Programmatic Access

\n
    \n
  1. Open AWS Console -> IAM -> Users -> Create user.
    2. \n
  2. Enable programmatic access.
    3. \n
  3. Attach required permissions.
    4. \n
\n

Recommended minimum approach:

\n\n

9.5 Create Access Keys

\n
    \n
  1. IAM -> User -> Security credentials.
    2. \n
  2. Create access key.
    3. \n
  3. Save AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY securely.
    4. \n
\n

9.6 Configure AWS CLI

\n
aws configure
\n

Set:

\n\n

9.7 Optional: Connect to S3 for Artifacts

\n

If you want to upload logs/reports:

\n
import boto3\n\ns3 = boto3.client(\"s3\", region_name=\"us-east-1\")\ns3.upload_file(\"results/sample_report.html\", \"your-bucket-name\", \"reports/sample_report.html\")
\n

9.8 Run Scripts Locally

\n
python main.py\npython demo.py\npython scenarios/login_scenario.py\npython scenarios/checkout_scenario.py
\n

10. Example SDK Usage

\n
from nova_act import NovaAct\n\nwith NovaAct(starting_page=\"https://example.com\") as nova:\n    nova.act(\"Login with valid credentials\")
\n

11. My Implementation

\n

I implemented this project as scenario-based automation to mirror realistic ecommerce testing flows. Each scenario is isolated for focused validation, and there is also a comprehensive scenario for end-to-end coverage.

\n

Implemented scenarios:

\n\n

This layout makes it easier to run individual suites or combine them for broader regression execution.

\n

12. Project Structure

\n
AI-Test-Automation-Using-Amazon-Nova-Act/\n  scenarios/\n    add_product_scenario.py\n    checkout_scenario.py\n    comprehensive_scenario.py\n    error_handling_scenario.py\n    forgot_password_scenario.py\n    login_scenario.py\n    login_with_report.py\n    product_count_scenario.py\n    product_count_with_report.py\n    registration_scenario.py\n    search_filter_scenario.py\n    wishlist_profile_scenario.py\n  results/\n  demo.py\n  main.py\n  test_result_logger.py\n  HTML_REPORTING_GUIDE.md\n  README.md\n
\n

13. Test Reporting

\n

The project includes HTML reporting for execution tracking.

\n\n

Typical report data includes:

\n\n

14. Benefits of AI Automation

\n

From an SDET perspective, key benefits are:

\n\n

15. Future Improvements

\n

Planned evolution areas:

\n\n

16. Author

\n

Aayush Mishra
\nSDET | Automation Engineer | AI Testing Enthusiast

\n

GitHub: https://github.com/Aayush-Mishraa
\nLinkedIn: https://www.linkedin.com/in/aayushmishra33/
\nPortfolio: https://aayushmishra.tech

\n
","tree":{"fileCount":21,"dirCount":1,"truncated":false,"extensions":[{"ext":"py","count":15},{"ext":"md","count":2},{"ext":"example","count":1},{"ext":"gitignore","count":1},{"ext":"docx","count":1},{"ext":"txt","count":1}],"topLevel":[{"name":"scenarios","type":"dir"},{"name":".env.example","type":"file"},{"name":".gitignore","type":"file"},{"name":"demo.py","type":"file"},{"name":"HTML_REPORTING_GUIDE.md","type":"file"},{"name":"main.py","type":"file"},{"name":"opencart_prd_aayush_mishra (1).docx","type":"file"},{"name":"README.md","type":"file"},{"name":"requirements.txt","type":"file"},{"name":"test_result_logger.py","type":"file"}]}}