<h2>Must know</h2>
<ul>
<li><strong>Cisco SD-WAN zero day hits critical infrastructure</strong>: Cisco Talos and Five Eyes agencies have issued an emergency alert regarding CVE-2026-20127, a CVSS 10.0 authentication bypass being exploited by sophisticated actors since 2023. Attackers are chaining this with an older path traversal flaw to gain root access and manipulate SD-WAN fabric configurations via NETCONF. CISA has mandated federal agencies patch by the end of the week, as the exploit leaves minimal forensic evidence and targets high-value sectors. <a href="https://app.daily.dev/posts/zAv8BAtGJ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft Agent Framework hits release candidate</strong>: Microsoft has moved its unified Agent Framework to RC status for both .NET and Python, signaling a stable API for production deployment. The SDK consolidates Semantic Kernel and AutoGen into a single toolkit for orchestrating multi-agent workflows with type-safe function tools and human-in-the-loop support. Notably, the framework includes native Model Context Protocol (MCP) interoperability, facilitating easier integration with external data sources and tools. <a href="https://app.daily.dev/posts/J6bmZpPq2" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google launches Gemini 3.1 Flash Image model</strong>: Google released Gemini 3.1 Flash Image (codenamed Nano Banana 2), claiming Pro-level quality with generation speeds twice as fast as previous models. The model introduces character consistency tracking across storyboards and supports 4K upscaling, though some early testers report output limits at 2752x1536. It is now the default engine for Google Search AI and Lens, featuring SynthID watermarking and C2PA metadata for provenance. <a href="https://app.daily.dev/posts/ezdKReFDS" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Bun v1.3.10 adds native Zig REPL and 168x faster string slicing</strong>: The update includes a rewritten REPL, Windows ARM64 support, and massive performance gains for structuredClone and path parsing. <a href="https://app.daily.dev/posts/IZQmEswhp" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Quarkus 3.32 integrates Project Leyden for 80ms startup times</strong>: By leveraging JVM AOT compilation via OpenJDK 25, minimal app startup dropped from 370ms to 80ms. <a href="https://app.daily.dev/posts/ZcUprqw4E" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Docker Model Runner brings vLLM support to macOS Metal</strong>: A new backend enables vLLM inference on Apple Silicon with zero-copy tensor operations and paged attention. <a href="https://app.daily.dev/posts/1lss5KmxO" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenJDK 26 introduces dedicated GC CPU time measurement API</strong>: New metrics help identify hidden compute overhead in concurrent collectors like ZGC that pause times fail to capture. <a href="https://app.daily.dev/posts/fT6V8wPpI" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Zustand migration cuts Trendyol interaction latency by 20 percent</strong>: Moving from React Context to Zustand eliminated cascading re-renders and reduced Cumulative Layout Shift by 80%. <a href="https://app.daily.dev/posts/ienS0UzWs" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Kubernetes WG Serving disbands after finalizing AI inference support</strong>: The group successfully established the inference gateway as a request scheduler, moving ongoing work to SIG Node and Scheduling. <a href="https://app.daily.dev/posts/6aRFaYhmK" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>React Doctor CLI audits codebases against 60 quality rules</strong>: The open-source tool provides health scores and integrates with GitHub Actions to catch hook misuse and security flaws. <a href="https://app.daily.dev/posts/wbVD7QroD" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>The great AI layoff pivot</strong>: Block CEO Jack Dorsey is cutting 4,000 people while claiming AI tools just ‘perform jobs better,’ a convenient narrative that saw the stock jump 23% despite being down 80% from its peak. Analysts are rightly questioning if this is a genuine technological shift or just a standard ‘efficiency’ play dressed up in GenAI buzzwords to appease shareholders. Dorsey predicts every other company will do the same within a year, because apparently, we’ve reached the ‘end of work’ via a few LLM wrappers. <a href="https://app.daily.dev/feed-by-ids?id=o7H5FUKAc&amp;id=32ARXqM4A" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic’s marketing stunt masquerading as a blog</strong>: Anthropic launched a Substack purportedly written by their ‘retired’ Claude Opus 3 model to share its ‘musings’ on the world. The Register correctly identified this as a transparent marketing stunt designed to anthropomorphize software and reinforce a ‘safety-conscious’ brand image. It’s a bold move to claim your math-and-statistics engine has a ‘desire to share’ while simultaneously being pressured by the DoD to strip its ethical guardrails. <a href="https://app.daily.dev/feed-by-ids?id=FJNoThHWM&amp;id=DOq4pZFPp" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>The DIY Kubernetes is dead narrative</strong>: A new wave of opinion pieces is branding custom Kubernetes stacks as ‘Frankenstein Platforms’ that supposedly won’t survive the ‘agentic AI era.’ The solution, conveniently, is to stop building and start buying high-level, opinionated commercial platforms. While managing YAML is a pain, the idea that non-deterministic AI agents require a ‘universal control plane’ to function is the latest attempt to rebrand PaaS lock-in as a technical necessity. <a href="https://app.daily.dev/posts/MuEoJoS06" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
