<h2>Must know</h2>
<ul>
<li><strong>OpenAI Codex and JetBrains AI integration for developers</strong>: OpenAI Codex, powered by GPT-5.2-Codex, is now natively integrated into JetBrains IDEs like IntelliJ, PyCharm, and WebStorm, starting with version 2025.3. This allows developers to access an AI agent for tasks such as refactoring, code migrations, and security vulnerability detection directly within their IDE, as reported by JetBrains. Users can control AI autonomy and resource allocation, with authentication via JetBrains AI subscription, ChatGPT account, or OpenAI API key. This integration aims to streamline large-scale coding tasks and is available free for a limited time. <a href="https://app.daily.dev/posts/uAvdGTErT" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents pose new risks for IT operations and security</strong>: AI agents introduce non-deterministic actions into IT operations, increasing risk, according to an article on IT operations automation. This is compared to ‘cowboy’ sysadmins making unreproducible changes. While AI can assist with exploratory tasks like generating playbooks and Dockerfiles, direct modification of live production systems is deemed unsafe. The recommendation is for AI to propose changes to version-controlled artifacts, emphasizing human review, testing, and reproducibility before deployment. Similarly, ZEST Security introduced AI Sweeper Agents to identify truly exploitable vulnerabilities, claiming over 90% of high-critical vulnerabilities are not exploitable in context, dismissing 11 million vulnerabilities in six months for early adopters. This highlights the need for context-aware security but also the potential for over-reliance on AI claims without independent verification. <a href="https://app.daily.dev/feed-by-ids?id=U8PYVjc73&amp;id=LDTl5G8Jm" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Ireland proposes expanded police surveillance with spyware</strong>: Ireland has introduced the Communications (Interception and Lawful Access) Bill, updating its 1993 surveillance framework to allow law enforcement to deploy commercial spyware for intercepting encrypted digital communications. This move comes amid a rise in spyware use across Europe, with judicial oversight and privacy safeguards promised. However, the articles do not provide specifics on implementation or independent oversight, raising concerns about potential abuses, similar to those documented in other European countries. <a href="https://app.daily.dev/posts/wO3SoInHG" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>GitHub releases multi-platform SDK for Copilot agent integration</strong>: GitHub has launched a multi-platform SDK for Python, TypeScript, Go, and .NET, enabling integration of Copilot’s agentic workflows, including orchestration, planning, and programmatic file edits, into various applications. <a href="https://app.daily.dev/feed-by-ids?id=GXXnGjqMp&amp;id=GuG7K2ryZ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Datadog launches automatic retention filters for RUM and APM data</strong>: Datadog released cross-product retention filters that automatically retain Application Performance Monitoring (APM) traces linked to Real User Monitoring (RUM) sessions, unifying frontend and backend observability data for improved troubleshooting. <a href="https://app.daily.dev/posts/fs5040Sq6" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>ClickHouse launches managed Postgres service with native integration</strong>: ClickHouse unveiled a new managed Postgres service, promising up to 10X faster performance for disk-bound workloads using NVMe storage, with native integration via Change Data Capture (CDC) and the pg_clickhouse extension. <a href="https://app.daily.dev/posts/WHTgaLabX" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Feast integrates with PyTorch ecosystem for feature management</strong>: Feast, an open-source feature store, has joined the PyTorch Ecosystem to provide a unified API for consistent feature management across model training and deployment, addressing training-serving skew in production AI systems. <a href="https://app.daily.dev/posts/dPG1hvjH4" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Rust 1.93.0 released with musl update and API stabilizations</strong>: Rust 1.93.0 is out, updating the bundled musl libc to 1.2.5 for improved DNS resolution in static Linux binaries, stabilizing 23 APIs, and allowing conditional inline assembly with cfg attributes. <a href="https://app.daily.dev/feed-by-ids?id=wF6DW7Hp4&amp;id=O9g10PE7Q" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Vercel Sandboxes introduces filesystem snapshot feature</strong>: Vercel Sandboxes now supports filesystem snapshots, allowing users to capture and launch new sandbox environments from saved states, streamlining CI/CD workflows, testing, and rapid prototyping by eliminating repetitive setup tasks. <a href="https://app.daily.dev/posts/Uv3Zx9otG" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Amazon Neptune Analytics expands regional availability</strong>: Amazon Neptune Analytics, a serverless graph database service, has expanded to seven new AWS regions, enhancing global accessibility for AI application development and offering managed GraphRAG integration with Amazon Bedrock. <a href="https://app.daily.dev/posts/mMcdehYea" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI agents are coding for you, what could go wrong?</strong>: Apparently, Claude Code with Opus 4.5 can handle ‘most development work’ without extra frameworks, and tools like Agent Flywheel promise to turn a VPS into a 24/7 autonomous AI coding environment for a mere $440-656/month. Because who needs human developers when you can have AI churning out production code autonomously? I’m sure the security risks and data privacy implications of non-developers creating applications without understanding vulnerabilities will just magically sort themselves out, as one discussion aptly compared it to ‘giving guns to people who’ve never seen them before.’ <a href="https://app.daily.dev/feed-by-ids?id=oecQfT5G1&amp;id=xBP8EMs83&amp;id=RlSMmScC7" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>eBay bans AI shopping agents, because who needs efficiency?</strong>: eBay, in its infinite wisdom, has updated its User Agreement to explicitly ban third-party AI agents and chatbots from making purchases without permission. This is apparently a response to the ‘rise in agentic commerce tools’ like ChatGPT’s Instant Checkout. Because heaven forbid we have efficient, autonomous shopping experiences that bypass human review. I’m sure this isn’t just about maintaining control and extracting more fees, but a genuine concern for the sanctity of human-driven online shopping. <a href="https://app.daily.dev/posts/eNz11CwvB" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI will make you dumber, says MIT, but don’t worry, it’s just ‘cognitive debt’</strong>: A four-month MIT study found that ChatGPT users exhibited the ‘weakest brain connectivity patterns’ and ‘consistently underperformed on neural, linguistic, and behavioral measures’ when writing essays. They even felt less ownership of their work and had reduced cognitive engagement when writing without AI later. The researchers are calling this ‘cognitive debt.’ So, while AI promises to make us more productive, it seems it’s also making us collectively dumber. But hey, at least we’ll have more time to scroll through AI-generated content, right? <a href="https://app.daily.dev/posts/k6I32Qj23" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
