Welcome to the 255th edition of Coding Jag brought to you by TestMu AI!š
AI devs, itās happening! š„ OpenAI just dropped gpt-oss, their first open-weight model since GPT-2, and the possibilities are massive. Whether you're running models locally or scaling them in the cloud, the control is now in your hands. Oh, and itās all live on Azure AI Foundry and Windows AI Foundry.
But thatās just the headline. This edition is bursting with updates that push the limits of modern development: from Aspire 9.4ās blazing-fast CLI to real-time Google Search grounding in Firebase, Shopifyās interactive AI UI, and how Salesforce crushed CI/CD bottlenecks with AI.
Plus, deep dives into how Whatfix uses TestMu AI to cut deployment cycles by 70% and triple test coverage, smarter GitHub Actions, privacy tools, accessibility tips, and three canāt-miss testing conferences.
Letās jump in! š„
News
07 min read
devblogs.microsoft.com
š Whatās new in Aspire 9.4? Maddy Montaquila shares the biggest release yet with a blazing-fast CLI, interactive dashboards, and seamless AI dev using GitHub Models & Azure AI Foundry. From smart parameter prompts to external API modeling, this update is packed for modern app dev!
05 min readazure.microsoft.com
š„ What if you could run OpenAI models your way? Asha Sharma & Logan Iyer announce gpt-oss, OpenAIās first open-weight models since GPT-2, now on Azure & Windows AI Foundry! Run gptāossā120b in the cloud or gptāossā20b locally. Full control, real performance, endless possibilities!
04 min readwww.digitalocean.com
šļø Whatās new in DigitalOcean Managed Databases? Nicole Ghalwash shares H1 highlights, from PostgreSQL 17 & MongoDB v8 support to 30TB storage, Valkey caching, RBAC, Datadog integration & Kafka Schema Registry. A powerful leap forward for scalable, secure data infrastructure!
04 min readshopify.engineering
šļø Shopify launches MCP UI, a powerful upgrade to the Model Context Protocol! Liad Yosef reveals how this new spec lets AI agents deliver rich, interactive components, think product selectors, image carousels, and full cart flows, right inside chat. Designed for seamless integration, these components automatically adapt to your appās style with CSS and render-data support, ensuring brand consistency across every interaction.
06 min readspreadprivacy.com
š”ļø DuckDuckGo launches 7-day free trials for Privacy Pro, its all-in-one privacy suite with VPN, Personal Info Removal, and Identity Theft Restoration. Now available internationally, with upgrades like phishing protection, Siri Shortcuts, auto VPN snoozing, and a redesigned data removal dashboard. Total privacy, now easier than ever. Try it now!
AI
06 min readengineering.salesforce.com
āļø How Salesforce used AI to slash test failure resolution time by 30%! Manoj Vaitla reveals the TF Triage Agent, an AI-powered system handling 6M daily tests across 78B combos. Built in just 6 weeks using Cursor, it combines FAISS search, smart LLM triage, and real-time context to boost dev trust and kill productivity bottlenecks in CI/CD. Engineering automation at its finest!
05 min readfirebase.blog
š Firebase AI just got real-time smarts! Ankita Saxena & Daniel La Rocque unveil Grounding with Google Search, now live in Firebase AI Logic SDKs. Gemini can now fetch live web data in-app, unlocking accurate answers about breaking news, sports scores, market trends & more. Available on Android, iOS, Flutter, Web & Unity. Real-time AI, right from your app.
Automation
10 min readwww.testmu.ai
ā” Curious how Whatfix speeds up releases and scales QA? By replacing their legacy Selenium Grid with TestMu AIās Web Automation platform, Whatfix cut test execution from 20 hours to just 3, reduced deployment cycles by 70%, and expanded test coverage 3x. With smarter CI/CD integration, improved test stability, and zero infra headaches, TestMu AI became central to delivering faster, high-quality digital experiences.
10 min readwww.atlassian.com
š Atlassian levels up Rovo Chat with Deep Research, a powerful new AI feature that generates detailed, citation-backed reports in minutes. William Li & team reveal how multi-path retrieval, structured generation, and reasoning models like Claude 3.7 drive context-aware insights for complex queries. Auto-export to Confluence, real-time data from Jira & 3rd-party tools, this is AI knowledge search done right.
06 min readgithub.blog
š¤ Bring AI into your workflows! Kevin Lewis shows how to automate bug triage, changelogs & issue summaries using GitHub Models in GitHub Actions. Learn to integrate AI via inference steps, CLI extensions, and custom prompts, plus pro tips on secure permissions and multi-model support. Code, commit, and let AI do the grunt work.
21 min readwww.testmu.ai
š§ Wondering how ARIA can make your web apps more accessible? Saniya Gazala breaks down ARIA (Accessible Rich Internet Applications) attributes with practical examples and key use cases. You'll learn the when and why behind aria-label, aria-hidden, aria-live, and more, plus common pitfalls to avoid. Whether youāre improving screen reader support or building dynamic interfaces, this guide helps you use ARIA the right way.
05 min readwww.softwaretestingmagazine.com
š° Curious how AI is transforming mobile casino testing in Ontario? Software Testing Magazine dives into how AI stress-tests RNGs, detects payout biases in real-time, validates ultra-fast withdrawals, and simulates heavy traffic, all while meeting AGCO standards. Faster compliance, fairer games, smarter QA.
Tools
10 min readwww.amazon.science
š§© What does it take to teach machines how to reason? Byron Cook dives into the core challenges, mapping natural language to logic, defining truth, and scaling formal verification. Learn how Amazonās Bedrock Guardrails tackle these issues to build trustworthy AI. A must-read for anyone curious about the logic behind safe, explainable systems.
15 min readwww.testmu.ai
š Looking beyond Postman for your API testing toolkit in 2025? Poornima Pandey explores 15 top Postman alternatives, from AI-native solutions like TestMu AI KaneAI to dev-friendly CLI tools like HTTPie and cURL. Whether you're aiming for faster test creation, seamless collaboration, or enterprise-grade performance, there's something for every workflow. Itās a great read if youāre rethinking your API stack and need practical, well-researched options to match your teamās evolving needs.
Video & Podcast
31 minengineering.fb.com
šļø In this episode, Pascal Hartig hosts Sean B., Lauren G., and Jesse M., research scientists at Metaās EMG team, to explore the future of wrist-based human-computer interfaces. They dive into how surface electromyography (sEMG) can enable subtle hand-gesture control and discuss their work on solving generalization challenges in neuromotor interfaces, blending neuroscience with engineering to build tech that works for everyone. š§ Listen now!
13 minwww.youtube.com
š„ In this video by Alan Richardson, aka EvilTester (@eviltester), demonstrates how to use Notebook LM for fast, effective studying and professional development. He explores features like auto-generated summaries, timelines, FAQs, and audio overviews from YouTube videos and other sources. A custom prompt is also shared for generating better transcripts. Ideal for learners and testers looking to accelerate knowledge with LLM tools!
Events
05 minwww.testmu.ai
š TestĪ¼ Conference 2025 returns from August 19-21! Be part of a global online gathering of 50,000+ testers and developers for expert-led tutorials, interactive workshops, live challenges, and over $10,000 in prizes. Discover cutting-edge testing tools, gain insights from industry leaders, and sharpen your skills, all at no cost. Register today!
05 minwww.eventbrite.com
šļø Join Ryan Pettibone online on Friday, August 15, at 12:30 AM IST for a 50-minute session hosted by Fixstars on AI Performance Engineering. Discover how to accelerate AI workloads through GPU profiling, practical optimization techniques, and the powerful capabilities of Fixstars AIBooster. With real-world case studies and a live Q&A, this session is a must for infrastructure managers and AI tech leads looking to cut costs and boost performance.
05 minassociationforsoftwaretesting.org
š CAST 2025, hosted by the Association for Software Testing, is happening August 25-27 in Salt Lake City, Utah! This flagship conference unites global quality professionals for three days of expert keynotes, hands-on workshops, real-world challenges, and tool demos. With transparent pricing, career-boosting sessions, IEEE CEU/PDH credits, and a vibrant community vibe, itās more than a conference, itās a movement shaping the future of software testing. Register early for exclusive member perks and group discounts!