Please turn JavaScript on
Association for Software Testing icon

Association for Software Testing

Is this your feed? Claim it!

Publisher:  Unclaimed!
Message frequency:  1.16 / day

Message History


Read full story

Evaluating test outcomes with LLM judges introduces hidden vulnerabilities when renderers silently drop structured tool calls and models struggle with visual reasoning. This episode explores how engineering teams can implement rigorous input assurance obligations and evidence sealing to guarantee evaluation integrity. Learn how to stop your AI judge from rubber-stamping blank...


Read full story

This episode covers GitHub Copilot’s deprecation of legacy frontier models, Katalon’s AI Test Runner for executing natural-language manual test cases, and AWS WAF architectures for securing stateless Model Context Protocol endpoints without blocking automated test runners. We also examine the DEIA HarmonyOS crash repair study, which highlights why AI-generated patches still r...


Read full story

I woke up today feeling really good about this journey I'm on. When I look back at where I started, I barely recognize that guy.

Closer to three years ago now, on Father's Day of 2024, I lost a job I'd loved for eight years when the company closed its doors. Before that, going back to before the turn of the century, I'd never been out of work longer than a single day, ...


Read full story

AI coding assistants routinely generate CRUD endpoints that query records by primary key alone while pairing them with single-tenant automated tests that pass because the mock caller owns the target object. This episode covers how to detect and prevent Broken Object-Level Authorization (BOLA) using a three-gate Dual-Tenant Verification Harness. Testers will learn how to enfor...


Read full story