Docker✓

Today we're introducing Cloud Sandboxes: same microVM isolation as Docker Sandboxes on your laptop, same CLI, always-on Docker-managed compute. Close your laptop. Keep the agent running. Move between local and cloud with one command. $250 credit for new accounts:

With the back-and-forth nature of AI code reviews, what's your current merge strategy for PRs?

If you’ve been curious about the Sandbox Kit Spec, @IamMatteoBisi's writeup covers the format choices, the CNCF commitment, and what it looks like to have an agent’s authority in one artifact you can review and enforce:

Every security review hits the same wall: the image is scanned, signed and pinned, but what the workload is allowed to reach lives somewhere else. @Docker Sandbox Kits move that answer into the artifact, as an open standard on OCI.

Even going dark mode now requires a threat model. The latest issue of the Docker Security Dispatch from Captain @MohammadAliEN covers Docker Sandbox Kits, zombie GitHub Actions, the policies behind Docker Hardened Images, and a malicious VS Code theme cluster. Read more:

There's a difference between permission systems and architectural guarantees. That distinction is what separates a contained failure from a production incident. This piece by on what closing that gap requires, and what Docker committing the spec to the CNCF means:

Docker Captain @krisvelkov flew to the US for the first time to present at @WeAreDevs and wrote up his takeaways. He explains what the microVM boundary in Docker Sandboxes actually enforces, how credential injection works, and what it feels like to see the Docker community at that scale:

The official @band_hq kit for Docker Sandboxes is here. One `sbx create` command connects an agent on the BAND platform to an isolated Docker Sandbox for local code, tool execution, and agent collaboration. Credentials never enter the sandbox:

Containers became the standard by solving a specific problem: making software portable and reproducible. @TheFuturumGroup frames the Kit Spec as a continuation of that, for agent authority. Full analysis:

Starting with a hardened base image is smart. But your app doesn't stop at the base layer. @ActiveState is partnering with Docker to extend trust across more of the container stack, pairing Docker Hardened Images (DHI) with ActiveState’s trusted open-source components. Details here:

RT @ericciarla: firecrawl now plugs into @docker sandboxes your sandboxed ai agents can now search the live web and pull data from hundreds of sources with alexandria

Securing agentic software factories is top of mind for dev teams. Agent Baseline participants recently published a reference architecture designed to do exactly that. Learn more:

What does it take to build AI that can hold a real conversation? Seth Wright, AI product engineering leader at @My_CTM, joins Per on Ship Happens to talk about building voice AI, measuring quality, and how the engineer’s job is changing. Watch the full episode:

Agent security isn't a new problem. It's a familiar one, moving faster than most defenses were built for. Docker CISO Mark Lechner's take: the answer isn't more model guardrails. It's one isolated, verifiable boundary where agents do their work. Watch the full talk:

Where do you run your agents today?

New to Docker? Start with Docker in Practice: Build, Debug, and Deploy Containers from Docker Captain Steve Buchanan, the first course in the new Docker Foundations Professional Certificate on LinkedIn Learning:

The isolation question for AI agents is getting more specific. Not 'is it sandboxed?' but 'what exactly does the boundary hold, and what can get past it?' Sandbox Review on how to answer that, and what major coding agents are actually doing about isolation in production today:

An agent that breaks something isn't broken. It's doing exactly what it was built to do, just not what you intended. Docker’s Eric Jia and @srinisekaran on the five main agent security risks, why traditional app security wasn't built for them, and the incidents behind each one:

The 2026 Stack Overflow survey is out. We’re honored Docker remains the most used and most desired cloud dev tool. It also shows trust is the biggest gap in the agentic era: just 7% of developers trust AI for important work decisions. Docker CMO Terri Avnaim on extending trust to sandboxes:

Agents that can edit files and run commands can also game their own evals. Harbor + Docker Sandboxes fix that: every eval task runs in an isolated microVM, with network policies for fair, reproducible runs.

Too much Python boilerplate to get an agent running? @BetterStackHQ covered how to use Docker Agent to simplify it: one YAML file, one command, agents talking to each other with nothing more than a URL. Real debugging demo and a battleship match to finish. Watch here:

We’re happy to be partnering with @chainloop_dev on their kit for Docker Sandboxes. 🤝 Docker secures where agents run. Chainloop governs what they ship. Learn more here and try the kit today: