TestForge AI turns written requirements into a working, self-healing test suite — and lets an AI analyst explain every failure in plain English.
You paste in a requirement document. TestForge reads it and drafts a complete suite of Gherkin scenarios for you to review. Approved scenarios become real, executable Playwright tests automatically — TestForge derives the page-object model from your live application by scraping the actual DOM, picks the right selectors, and scaffolds the test files across Chromium and Firefox. You do not write the Playwright code, you do not write the page objects, you do not write the assertions.
Every regression run spins up a clean disposable browser container, captures full-page screenshots of every step, compares them pixel-by-pixel to your baseline, and produces a structured report. When something fails, an AI analyst — built on Anthropic's Claude — classifies the failure, explains it in plain English (Category, What was expected, What happened, Most likely causes, Suggested fix), and drafts the Jira ticket if escalation is warranted.
Self-healing locators recover from common UI changes — a renamed button, a shifted CSS class, a restructured section — using role, label, accessible name, and adjacent-element text. Over time TestForge learns the locator drift pattern and proposes permanent updates to your page-object model.
Every confirmation or dismissal you give the platform becomes training signal. The system you use in month six is not the same system you used in month one — it knows your application.
Built on five operationalized microservices (Test Case Generation, Test Prioritization, Self-Healing Automation, AI Governance, Defect Prediction). Underlying research is documented in six journal manuscripts at IEEE, Wiley, Elsevier, and Springer venues. Permanently archived datasets on Zenodo (DOI 10.5281/zenodo.19682732).
