#053 • 30 December 2025

Contents of this Newsletter

🎊

A Quiet Thought to Carry Into the New Year

⏰ : 4 min

✅

Quash Tests Your Apps

⏰ : 4 min

🔔

What We Shipped Last Week

⏰ : 2 min

📓

Blog Picks

⏰ : 10 min

🗞️

Weekly AI News

⏰ : 5 min

MAIN STORY

If You’re Reading This, Happy New Year!

If you’re reading this sometime between wrapping up the year and easing into the next one, happy new year. There’s something about the end of the year that makes everything feel a little slower. Not in a bad way, but it’s just more deliberate. Fewer grand plans, more reflection on what actually worked and what didn’t.

This is usually the part of the calendar where attention is scattered, inboxes are noisy and most plans are quietly postponed to “after things settle down.”

Instead of trying to say too much, we wanted to share one idea that stayed with us this year and feels worth carrying forward, especially for teams building and testing mobile apps.

Where Mobile Testing Still Breaks Down

Most teams today have made significant progress in running tests efficiently and achieving reliable execution is still a complex and important challenge that requires careful setup and maintenance. Over the last few years, improvements in speed, coverage and automation have made it possible to handle execution at scale, but solving for this foundational part of testing is only the first step.

Where things still fall apart is what happens after a test fails. When something breaks, teams are often left piecing together what actually happened by jumping between screenshots, logs, videos and assumptions. Even with plenty of tests in place, failures often arrive without enough context to make them actionable.

In mobile testing, this problem shows up even more clearly. State plays a much bigger role. Where you are in a flow, what actions happened before and why the screen looks the way it does all influence the outcome. When that context is missing, a failure does not just slow debugging down. It blocks progress entirely.

Adding more tests does not solve this. Making failures clearer does.

What Seeing This Pattern Changed for Us

Watching this play out across teams changed how we think about testing. We stopped looking at failures as a byproduct of execution and started treating them as the most important output of a test run.

If a failure cannot explain what went wrong and why it happened, then speed, scale, and coverage lose much of their value. The real cost shows up later, when engineers spend time reproducing issues, guessing at state, or trying to reconstruct what a test actually did. Those slowdowns rarely come from missing tests. They come from missing clarity.

What We’re Carrying Into the New Year

Going into the new year, that shift in perspective has helped us stay focused. We are carrying forward strong opinions about state and context, a bias toward failure clarity over raw execution speed and a commitment to building around real workflows rather than idealized demos.

At the same time, we are deliberately moving away from optimizing for volume in isolation, where adding more tests becomes a proxy for quality even when failures remain unclear. We are also stepping away from the idea that flaky outcomes are simply a cost teams have to absorb, or that it is acceptable for tools to provide less information precisely when something breaks. Scale and coverage still matter, but only when they lead to clearer understanding, because once you see how much time is lost to ambiguity, those trade-offs become difficult to justify.

We will have more to share as the year unfolds. For now, we hope you get a little distance from dashboards and test runs and come back into the new year thinking a bit more critically about what effective testing should actually look like.

Your Apps, We Test

Each week, we put our execution engine to work on real apps.

This week, we put our execution engine to work on Snitch, the fast-fashion menswear brand, to test its gift card purchase flow using QA mode. Using a natural language prompt, we asked Quash to open the app, navigate to the Gift Cards section, select a Happy Birthday card, choose a ₹2000 amount, fill in recipient details, and preview the card before purchase.

During execution, Quash validated the preview screen against the expected inputs and surfaced multiple issues in real time. The card title failed validation because the preview displayed “Congratulations” instead of “Happy Birthday.” While the amount value passed validation and was rendered correctly, the recipient details were missing from the preview altogether.

Running this flow in QA mode allowed us to catch content mismatches and missing data before checkout, turning what would normally be a visual spot-check into a structured, automated validation.

Want us to test your app next? Drop us a mail at hello@quashbugs.com

What Did We Ship?

Quash App

v1.1.4 | December 30, 2025

This week, we shipped a set of improvements across the Quash app.

Test reports are now generated faster and with improved reliability, making it easier to review results as soon as a run completes. We’ve also fixed issues related to test suite execution across multiple devices, ensuring more consistent behavior when running tests at scale.

Alongside these updates, we’ve rolled out several UI enhancements across the app to improve clarity, navigation, and overall usability.

These changes focus on making day-to-day testing smoother, faster and more dependable for teams running frequent or large test suites.

Check out our Recent Blogs

News

This Week In AI

Meta buys AI startup Manus to supercharge agent capabilities

Meta Platforms has acquired Manus, a Singapore-based AI agent startup known for autonomous, general-purpose agents that can execute multi-step tasks like research, coding, and workflow automation. The deal, valued at roughly $2 billion, will integrate Manus’s technology into Meta AI and other products while continuing to operate Manus as a standalone service. Manus has processed trillions of tokens, built millions of virtual machines, and reached rapid revenue milestones, making it a strategic addition to Meta’s AI push amid competition with OpenAI and Google.

SoftBank to Buy DigitalBridge for ~$4B to Boost AI Infrastructure

SoftBank Group has agreed to acquire DigitalBridge Group, a U.S. digital infrastructure investor specializing in data centers, fiber, and connectivity assets, in a deal valued at about $4 billion (~$16 per share in cash). The move is part of SoftBank’s strategy to deepen its footprint in AI-related infrastructure, securing capacity and connectivity needed to scale next-generation AI services. DigitalBridge will continue operating independently under its current leadership after the expected close in the second half of 2026.

Scientists Identify Two New MS Subtypes With AI

Researchers at University College London and Queen Square Analytics have used AI, blood tests, and MRI scans to uncover two distinct biological subtypes of multiple sclerosis—named early sNfL and late sNfL, that differ in how and when nerve damage and brain changes occur. Early sNfL shows high markers of nerve damage and rapid lesion formation, while late sNfL shows brain shrinkage first with later biochemical changes. The discovery could shift MS diagnosis from symptom-based categories to biology-driven treatment plans, enabling more personalized care and better outcomes for patients worldwide.

All Newsletter Issues | Security | Contact Us | Blogs

Before You Log Off For The Year 😴

Product

Solutions

By role

By use-case

By industry

Resources

Comparison

Company

Product

Solutions

By role

By use-case

By industry

Resources

Comparison

Company

Get Quash

Product

Solutions

By role

By use-case

Resources

Comparison

Company

Get Quash