#16 Digest
This week’s edition digs into practical fixes for tricky Playwright timing quirks, especially on Firefox, alongside fresh takes on AI-driven test workflows and smarter project structures. You’ll also find clever ways to shrink test infrastructure and insights on blending human QA with autonomous automation for smoother releases. Dive in for tools and tips that’ll help you ship faster and with more confidence.
The Scaffold: Playwright Project Structure Built for AI
This article introduces a Playwright project scaffold designed for AI-assisted test automation, detailing its structured folder setup, coding conventions, and integration with AI tools to improve test development efficiency.
AI Regression Tests Written in Markdown, Not Code
This article presents a novel approach to AI-driven regression testing using markdown files interpreted by an AI agent that interacts with a web app through accessibility trees rather than brittle selectors, complementing Playwright E2E tests by focusing on user-centric, resilient testing.
Human-in-the-Loop Testing: Best Practices to Combine AI and Human QA
This detailed article explores best practices for combining AI and human QA in testing workflows, with practical insights on using Playwright for human-in-the-loop testing, automated failure clustering, flaky test detection, and maintaining human control for test intent and review.
AI Browser Automation: 5 Layers Every Agent Builder Should Know
This article provides a detailed breakdown of AI browser automation across five layers, highlighting how Playwright and related frameworks integrate AI to improve test resilience, reduce maintenance, and automate workflows with semantic understanding rather than brittle selectors.
Why Vercel's agent-browser Is Winning the Token Efficiency War for AI Browser Automation
This article compares Vercel's agent-browser CLI tool against Playwright MCP for AI browser automation, highlighting agent-browser's superior token efficiency, architecture, and use cases while noting limitations and recommending when to use each tool.
My MCP Server Was 2x Larger Than Playwright. Now It's 136x Smaller.
This article describes the development and optimization of Charlotte, an open-source MCP server compared to Playwright’s MCP server, focusing on reducing serialization bloat and improving token efficiency for AI agents interacting with web pages.
How We Fixed Firefox's localStorage Race in Playwright: Two Navigation Helpers
This article discusses a Firefox-specific timing issue when using Playwright's addInitScript to seed localStorage, and presents a practical solution by splitting navigation helpers to improve test reliability and speed.
The Server-Side Mocking Gap Nobody Talks About
This article details a technical approach to server-side mocking for Playwright tests in modern fullstack applications using TanStack Start and MSW, addressing challenges of mocking server-originated requests with per-test scoped mocks and integrating them into E2E workflows.
I Shipped 126 Tests Last Month. Here's the AI Workflow That Got Me There.
This article details an AI-powered workflow using skill files and agents to efficiently generate comprehensive API and Playwright UI tests, including setup, patterns, and quality gates to improve productivity in test automation.
Solved: Are any codeless test automation tools worth using?
This article provides a detailed and critical analysis of codeless test automation tools, offering practical advice on when to use them versus adopting code-based frameworks like Playwright for scalable and maintainable testing.
Zero-Downtime Autonomous Deployments: AI-Driven QA for Multi-Site CMS
This article presents an autonomous QA pipeline leveraging Playwright for automated visual and functional testing across multiple regional sites, integrating with CI/CD workflows to reduce manual QA bottlenecks and ensure deployment reliability.
How Juvare restored test reporting visibility during their Cypress-to-Playwright
This article details Juvare's migration from Cypress to Playwright and how they restored centralized test reporting using Currents, providing real-time insights that improved test stability and troubleshooting across multiple projects.