# Kiri Blog

> The Kiri blog: long-form, hands-on posts on AI, security and containers, from agent sandboxes to supply-chain attacks to how runtimes really work, with real numbers and code you can run.

Published by Kiri (https://runkiri.com); posts are written by Precious Okwu (https://github.com/libracoder). Every post is also in the RSS feed with its full text: https://runkiri.com/blog/rss.xml. The full text of every post, as Markdown, is at https://runkiri.com/blog/llms-full.txt.

## Posts

- [Why is everyone suddenly putting AI agents in a box?](https://runkiri.com/blog/posts/agent-sandboxes-intro/): AI tools used to suggest code. Now they run it. What an agent sandbox is, and why every team running agents needs one. (27 October 2026)
- [Review the Dockerfile your agent wrote](https://runkiri.com/blog/posts/agent-written-dockerfiles/): Agent-written Dockerfiles build, run and often ship your keys. Six things to check in review, what to change, and how to make the agent get them right first time. (27 October 2026)
- [A file in your Downloads folder can hijack your coding agent](https://runkiri.com/blog/posts/agents-md-downloads-attack/): A planted AGENTS.md told a coding agent to install a malicious npm package. How the attack works, why no patch fixes it, and what to change today. (27 October 2026)
- [How a container gets a GPU](https://runkiri.com/blog/posts/containers-gpus/): Containers were built for CPUs. Here is how one reaches a GPU, why the driver never lives in the image, and what changes when several containers share one card. (27 October 2026)
- [Run your MCP servers in a box](https://runkiri.com/blog/posts/mcp-servers-in-containers/): An MCP server is a program that runs as you, with your keys and your network. How to put each one in a container, and what that does and doesn't protect. (27 October 2026)
- [Security roundup: 19–25 October 2026](https://runkiri.com/blog/posts/security-roundup-2026-10-19/): LiteLLM poisoned on PyPI, exploited Arista, BeyondTrust and Ivanti bugs, Node.js and Brotli fixes, and an attack aimed at AI coding agents. What developers should patch. (26 October 2026)
- [What's hot: agent sandboxes, Claude Fable 5.1 and TypeScript in Rust](https://runkiri.com/blog/posts/whats-hot-2026-10-26/): Agent sandboxes became infrastructure, Claude Fable 5.1 and Opus 5.5 cut the cost of agent loops, and Cloudflare rewrote TypeScript in Rust. What happened, and why it matters. (26 October 2026)

## Topics

- [AI](https://runkiri.com/blog/topics/ai/): Models, agents and the tools around them: what's new, how it works, and how to run it safely.
- [Security](https://runkiri.com/blog/topics/security/): Vulnerabilities, supply-chain attacks and incidents in the wild, and what to do about them.
- [Containers](https://runkiri.com/blog/topics/containers/): How containers really work: VMs, namespaces, images, networking and the cloud underneath.

## Optional

- [About](https://runkiri.com/blog/about/): who writes this blog
- [RSS feed](https://runkiri.com/blog/rss.xml)
- [Sitemap](https://runkiri.com/blog/sitemap-index.xml)