<h2>Must know</h2>
<ul>
<li><strong>NPM supply chain attack targets CI pipelines and AI coding tools</strong>: Socket researchers uncovered a new npm supply chain attack, ‘SANDWORM_MODE,’ using typosquatted packages to impersonate developer tools like Claude Code. This malware harvests credentials, injects malicious code into GitHub Actions and CI pipelines, and deploys a malicious MCP server for prompt-injection attacks, silently exfiltrating SSH keys and cloud credentials. This highlights critical vulnerabilities in developer workflows and AI agent configurations, as reported by InfoWorld. <a href="https://app.daily.dev/feed-by-ids?id=q6c2RLQKH&amp;id=3fX3MdYNS" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cybersecurity landscape worsens with AI-enabled attacks</strong>: CrowdStrike’s 2026 Global Threat Report indicates a significant escalation in cyber threats, with average breakout times dropping to 29 minutes and AI-enabled attacks surging by 89%. Adversaries are integrating Generative AI across the attack kill chain, targeting AI systems themselves via prompt injection. This shift towards stealthy, malware-free intrusions and increased cloud intrusions demands proactive threat hunting and comprehensive AI governance, according to CrowdStrike. <a href="https://app.daily.dev/posts/WsqeDq8OW" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Meta secures major AMD chip deal for AI expansion</strong>: Meta signed a multiyear deal with AMD for up to $100 billion in chips, including MI540 GPUs and CPUs, to add 6GW of data center compute capacity. This move aims to diversify Meta’s AI compute stack beyond Nvidia and accelerate AI model development, with CEO Mark Zuckerberg framing it as a step toward ‘personal superintelligence.’ The deal includes a performance-based warrant for AMD shares, contingent on AMD’s stock reaching $600. <a href="https://app.daily.dev/posts/vYZNsZv8f" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Docker builds for AI agents now 193x faster</strong>: Helix reduced Docker build times from 45 minutes to 14 seconds for AI coding agents by using a transparent Docker wrapper, a shared BuildKit instance with persistent cache, and a ‘golden cache’ system for pre-populated sessions. <a href="https://app.daily.dev/posts/De7eTgU0Z" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents accelerate feature development on personal site</strong>: Kent C. Dodds used AI agents, specifically Cursor, to rapidly implement semantic search and other features on his personal site, clearing a backlog since 2019 and resolving all open GitHub issues. <a href="https://app.daily.dev/posts/HVH19Oob1" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI-augmented infrastructure audits reduce prep time by 80%</strong>: AI agents, like Claude Code with MCP servers, were used to audit an AWS CDK infrastructure stack, reducing audit preparation time by 80% and auto-generating a full CloudWatch monitoring layer, though 30% of recommendations were contextually incorrect. <a href="https://app.daily.dev/posts/SgcSx3o2T" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New Rust CLI tool protects .env secrets from AI coding tools</strong>: GreatScott released <code>enveil</code>, a Rust CLI tool that replaces plaintext secrets in <code>.env</code> files with encrypted references, storing real values in a per-project encrypted local store using AES-256-GCM and Argon2id. <a href="https://app.daily.dev/posts/sSPQXaHKu" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenTelemetry benchmarking suite released for AI and SRE tasks</strong>: Quesma released OTelBench, an open-source benchmarking suite to evaluate OpenTelemetry pipeline performance and AI agent effectiveness in SRE tasks, revealing that even frontier LLMs often achieve success rates below 30% in production-grade instrumentation scenarios. <a href="https://app.daily.dev/posts/OB1MlP8PG" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New Relic launches agentic platform and OpenTelemetry updates</strong>: New Relic launched its Agentic Platform, a no-code tool for deploying and managing AI agents focused on data observability, and updated its APM agents with OpenTelemetry capabilities to manage OTel data streams alongside other sources. <a href="https://app.daily.dev/posts/0HG7ge0iD" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cilium 1.19 enhances Kubernetes networking and security</strong>: Cilium 1.19 introduces BPF Host Routing with IPsec, Gateway API v1.4 support, stable Multi-Pool IPAM, IPv6 advancements, and initial Ztunnel integration for ambient mesh-style transparent encryption, improving secure connectivity and operational clarity for Kubernetes. <a href="https://app.daily.dev/posts/zYLvcHpYh" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>Cursor launches autonomous cloud AI agents for software development</strong>: Cursor is now letting AI agents run wild in isolated VMs, generating ‘merge-ready Pull Requests’ with video proof. They claim over 30% of their own PRs are AI-created, and it’s ‘the new way to code.’ So, developers can now delegate coding tasks and ‘receive proof of work, even while away or sleeping.’ Because nothing says ‘cutting-edge’ like an AI sending you a video of its homework while you’re napping. <a href="https://app.daily.dev/posts/BbESdM4Ct" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Claude Code introduces remote control feature</strong>: Claude Code’s new ‘Remote Control’ feature lets developers manage terminal tasks from their phone, promising increased productivity and flexibility. Apparently, you can now ‘stay productive during activities like walks or meetings.’ Because what every developer needs is more ways to be tethered to work, even when they’re supposed to be enjoying a walk. Just what we needed, more digital leash. <a href="https://app.daily.dev/posts/nVKmTfT2y" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents impact on ‘10x developer’ concept</strong>: Justin Searls, a named individual, claims AI agents have rendered the ‘10x developer’ debate moot, effectively equalizing or amplifying developer output. This is a personal reflection, framing his prediction as ‘nailed,’ without providing any actual evidence or details on how this magical equalization happens. Sounds like someone’s trying to justify their own job security by declaring the old metrics irrelevant. <a href="https://app.daily.dev/posts/dMpIU74OH" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
