Testing GSD: From a Docs-Only Repo to Working Go Code in One Session

Introduction I have been experimenting with Spec-Driven Development for a while now. If you are not familiar with the approach, I have a few articles tagged spec-kit that cover the theory and a real hands-on walkthrough where I built a Go TUI for Apple Container management. The short version: instead of vibe-coding with an LLM and hoping for the best, you invest upfront in a structured specification, then let the AI work against that spec. The results are measurably different. ...

April 13, 2026 · 19 min · 3968 words · Matteo Bisi

Docker Sandboxes: Running AI Agents in YOLO Mode, Safely

A few days ago, Docker published an article on LinkedIn about a new tool called Docker Sandboxes (sbx). The pitch is simple: run AI coding agents in fully autonomous mode, without worrying about them touching your host machine. I read it and decided to install it on my MacBook Pro M4 (32 GB RAM) and test it for real. Not to read the documentation and summarize it, but to actually break things, observe what happens, and verify the security claims hands-on. ...

April 7, 2026 · 17 min · 3567 words · Matteo Bisi

Hardening ACTUI: Dependabot and OpenSSF Scorecard for a Side Project

The Unexpected Swag from KubeCon EU 2026 KubeCon EU 2026 Amsterdam was a great edition. I walked away with good conversations, new connections, and the usual conference bag full of stickers. But one thing stood out among the swag: six months of GitHub Copilot Pro+, courtesy of GitHub. I’m not going to pretend I wasn’t excited. Copilot Pro+ isn’t cheap, and having it handed to you as conference loot—just because you showed up in the right place, accepting the right invitation—felt like a proper thank-you to the community. GitHub clearly knows its audience. ...

April 2, 2026 · 8 min · 1619 words · Matteo Bisi

The Trivy Supply Chain Attack: A Breakdown of Credential Theft and the CanisterWorm Escalation

Introduction Trivy, the widely adopted open-source security scanner from Aqua Security, is a cornerstone of modern CI/CD pipelines and container security. With over 33,000 stars on GitHub as of March 2026, its footprint spans across Docker images, Homebrew, and countless developer machines. This ubiquity, however, made the supply-chain compromise discovered between March 19–21, 2026, particularly devastating. The incident was not a single point of failure but a multi-stage attack involving malicious releases, manipulated GitHub Actions, and a self-propagating worm that leveraged decentralized infrastructure. ...

March 21, 2026 · 4 min · 736 words · Matteo Bisi

GitHub Copilot: The High-ROI Multi-Model Powerhouse

Next week, I’ll be in Amsterdam for KubeCon EU 2026 (you can read my preview here), and my journey starts this Monday with the GitHub Social Club: Amsterdam. I was lucky enough to snag an invitation via LinkedIn and jumped at the chance to join. As someone who has used GitHub for years (both for personal projects and corporate needs) I’ve always appreciated the platform’s reliability. However, after a month of putting GitHub Copilot Pro through its paces, I’m genuinely surprised it isn’t even more ubiquitous. If you aren’t a “super heavy” coder but want access to the best tools, this is arguably the highest value-for-money platform in the AI space right now. ...

March 17, 2026 · 4 min · 704 words · Matteo Bisi

August 2026 Countdown: Are Your K8s AI Workloads EU AI Act Ready?

The countdown is on. As of March 2026, we are less than five months away from the full applicability of the EU AI Act (August 2, 2026). For those of us running AI workloads on Kubernetes, whether it’s self-hosted inference engines like vLLM or RAG-based agentic systems, compliance is no longer a legal “later” problem. It’s an engineering “now” problem. Before we dive into the technical implementation, let’s look at the critical roadmap and the cost of getting it wrong: ...

March 16, 2026 · 5 min · 1040 words · Matteo Bisi

The Exploitability Gap: Insights from Datadog’s State of DevSecOps 2026

Intro We have all been there: a Slack notification triggers an alert for a “Critical” CVE, and the scramble to patch begins. But as our clusters grow, so does the noise. The most jarring security stories are often the ones happening silently inside our own production environments. Datadog recently released its State of DevSecOps 2026 report, and the numbers provide a sobering reality check for anyone managing cloud-native infrastructure. The report reveals that 87% of organizations are currently running at least one known exploitable vulnerability in their deployed services. Even more concerning is that many of these services rely on libraries that have been abandoned by their maintainers. This is not just a theoretical problem; it is based on telemetry from thousands of real-world cloud environments, making the findings impossible to dismiss. ...

March 6, 2026 · 3 min · 626 words · Matteo Bisi

ACTUI Follow-Up: Submenus and Image Management

Quick Follow-Up After publishing the initial ACTUI article, I kept developing the tool. I started using it regularly and shared it with my team. Some feedback came in, and I naturally improved things during my free time. This is a quick update on what changed. What Changed Submenu Structure The original flat menu worked for a demo but felt cluttered with more features. I restructured the interface into three main sections: ...

February 27, 2026 · 2 min · 355 words · Matteo Bisi

Back to Basics: Why Containers Are Just Fancy Linux Processes

The path into platform engineering has changed. Many engineers today start their careers working directly with Kubernetes, writing YAML and managing Helm charts before they ever spend extended time at a Linux terminal. The tooling is so well-abstracted that you can be genuinely productive for months before the underlying system ever becomes relevant. That is a real achievement for the ecosystem. The gap shows up at the worst moments, though: a container crashes with a permission error, a security team flags a pod running as root, a privilege escalation CVE lands and it is not clear whether the cluster is exposed. These are Linux problems, and they are much easier to reason about once you understand what the YAML actually maps to at the kernel level. I have been in those conversations many times, and I always come back to the same set of fundamentals. ...

February 20, 2026 · 11 min · 2292 words · Matteo Bisi

The Challenge of Securing AI Agents: A DevSecOps Perspective

As a DevSecOps Team Leader, my job is to secure customers using modern technologies. Sounds straightforward, right? The reality is far more complex. Every day, I face the challenge of enabling innovation while maintaining security. The rapid adoption of AI has introduced a new dimension to this challenge: agentic AI assistants that do not just chat, they act. This challenge connects directly to something I wrote about recently. In my article on spec-driven development with GitHub Spec-Kit, I discussed how structure and governance matter when using AI for coding. The same principle applies here: when AI agents can execute code, access secrets, and operate with user privileges, we need structure and governance more than ever. ...

February 17, 2026 · 5 min · 1059 words · Matteo Bisi