CODING JAG - Issue 251

Welcome to the 251st edition of Coding Jag brought to you by TestMu AI!👐

Large Language Models (LLMs) are transforming how you work, but they also introduce serious security risks. One of the major ones is prompt injection, where attackers can manipulate the model to behave in unintended ways.

In this edition, learn how to catch real attacks with Python and GPT-4o-mini, explore Veo 3’s cinematic AI, Oracle 23.7 dev tools, GitHub Copilot customization, agentic workflows, UX pitfalls, AI agent showdown and more!

Let’s get into it 👇

News

Veo 3 Available for Everyone in Public Preview on Vertex AI

03 min readChrome-Extensioncloud.google.com

🤷🏻‍♀️ What if AI could create cinematic videos with synced audio, motion, and emotion, from a single prompt? Jason Gelman unveils Veo 3, now in public preview on Vertex AI, bringing near-Hollywood quality video generation to your fingertips.

Introducing Deep Research in Azure AI Foundry Agent Service

04 min readChrome-Extensionazure.microsoft.com

🥴 Curious how enterprises are scaling research with precision and transparency? Yina Arenas introduces Deep Research in Azure AI Foundry Agent Service, an API+SDK for building composable, auditable research agents using OpenAI’s o3-deep-research model, Bing Search, Logic Apps and more.

What’s New for Developers in Oracle Database 23.7

14 min readChrome-Extensionblogs.oracle.com

💻 Oracle Database 23.7 brings powerful new tools for developers! Gerald Venzl introduces features like the TIME_BUCKET function for custom time grouping, materialized columns for faster reads, and DBMS_DEVELOPER.GET_METADATA() for blazing-fast JSON metadata, JavaScript-to-PL/SQL FFI, and Smallfile tablespace shrinking to recover space.

Introducing the Awesome GitHub Copilot Customizations Repo

05 min readChrome-Extensiondevblogs.microsoft.com

🔧 Personalize GitHub Copilot like never before! Matt Soucoup and Aaron Powell unveil the Awesome GitHub Copilot Customizations repo, a community-powered collection of custom instructions, reusable prompts, and chat modes.

AI

A Practical Guide to Building Agentic AI for Enterprise Workflow Efficiency

10 min readChrome-Extensionblog.kore.ai

😁Want to build autonomous, intelligent workflows in your enterprise? Abhijit Mhetre breaks down Agentic AI, goal-driven systems that make decisions and take action with minimal human input. Learn the key components, use cases like invoice processing & recruitment, and best practices to scale efficiently.

How to Build Custom AI Agents With NVIDIA NeMo Agent Toolkit Open Source Library

04 min readChrome-Extensiondeveloper.nvidia.com

🤔 Wondering how AI agents are evolving from task bots to autonomous decision-makers? Nicola Sessions and Matthew Penn unpack how Agentic AI shifts workflows from rule-based automation to intelligent, goal-driven agents. Explore key components, use cases like claims and recruitment, and real-world best practices.

Automation

Snap One Accelerates Test Execution by 50% After Switching to TestMu AI

10 min readChrome-Extensionwww.testmu.ai

🎯 Discover how Snap One, a smart home and connected living solutions provider, transformed its QA workflow by switching from a flaky legacy setup to TestMu AI’s robust, AI-native testing platform. Struggling with unstable E2E tests, BDD integration gaps, and slow cross-browser execution, the team embraced TestMu AI’s HyperExecute and Test Insights to drive real results.

Testing LLMs for Prompt Injection Vulnerabilities

11 min readChrome-Extensionwww.thegreenreport.blog

🚨 Prompt injection is no longer just a theoretical threat; it’s actively targeting LLM-powered apps. Irfan Mujagić shares how to automate prompt injection testing using Python and OpenAI’s GPT-4o-mini. Learn how QA teams can detect jailbreaks, suffix attacks, and indirect injections before they hit production.

Everything to Know to Start Web Scraping in Python Today

10 min readChrome-Extensionscrapfly.io

📦 Overwhelmed by the maze of modern web scraping? Bernardas Ališauskas breaks it all down in this all-in-one Python guide, from basic HTML scraping to handling JavaScript-heavy pages using headless browsers like Selenium & Playwright. Learn to parse dynamic content, detect hidden datasets, and scale with AI-driven tools.

Simplify Development With Public Ports in Firebase Studio

05 min readChrome-Extensionfirebase.blog

😮‍💨 Tired of juggling backend access in full-stack dev? Chandra Sekhar Pydi shares how Firebase Studio’s public ports simplify integration between your frontend and backend - no auth, no code changes! Preview apps publicly, fetch APIs directly, and share feedback in seconds.

Advanced Authentication and Onboarding Workflows With Docusign Extension Apps

08 min readChrome-Extensionwww.thisdot.co

🤩 Excited to see how real-world onboarding challenges are tackled by extending OAuth 2 in Docusign Extension Apps? Dario Djuric demonstrates how to integrate systems like TaskVibe, without OAuth 2 support, by securely collecting and storing API keys.

Tools

The Best JavaScript Web Scraping Libraries

11 min readChrome-Extensionwww.scrapingbee.com

😣 Choosing the right JavaScript tool for web scraping in 2025? Ilya Krukowski compares top libraries like Playwright, Puppeteer, Cheerio, JSDOM, and more. From handling simple static sites to scraping dynamic JS-heavy pages, explore code samples, pros, and cons of 12+ tools, including SaaS platforms like ScrapingBee!

15 Best Project Management Software Tools for 2025

22 min readChrome-Extensionmonday.com

✨ Overwhelmed by scattered tools and messy workflows? Victoria Landsmann explores 15 top project management tools for 2025, from visual boards in Trello to AI-powered automations in monday work management. Discover trends, features, and expert tips to help your team deliver smarter, faster, and more securely.

Video & Podcast

Podcast: Bad Quality By Design

28 minChrome-Extensiontestingpeers.com

🎙️ In this episode of the Testing Peers podcast, David, Chris, Sanne, and Beth discuss Bad Quality By Design, how products are intentionally made worse over time to boost profits. They explore deceptive UX patterns, forced upsells, frustrating cancellation flows, and the loss of pride in product quality. Listen now!

Video: AI Agent vs Agentic Agents - Strength and Weakness

19 minChrome-Extensionwww.youtube.com

🎥 In this video by ExecuteAutomation, you'll explore the battle between traditional AI agents and emerging agentic agents. Find out which handles context better, whether multi-agent systems offer real autonomy, and how context compression and memory affect performance. Watch now!

Events

Testμ Conference 2025

05 minChrome-Extensionwww.testmu.ai

📅 Join Testμ 2025, happening from August 19th to 21st. Connect with a global network of 50,000+ professionals and participate in exciting challenges, attend workshops and tutorials featuring the latest tools and technologies, learn from expert sessions and keynote speakers, and stand a chance to win prizes worth up to $10k+. Don't miss out on the ultimate virtual conference for testers and developers! Register now!

Newcastle Meetup

05 minChrome-Extensionwww.ministryoftesting.com

📅 Join us for a hands-on BDD Workshop on 15 July 2025, from 18:00-20:00 BST at ThoughtWorks, Newcastle upon Tyne! Led by Peter Simons, this interactive session explores real-world Behaviour-Driven Development practices to improve software quality and collaboration.