k7

Not just the LLM, but any code that the LLM outputs also has to be firewalled. Sandboxing your LLM but then executing whatever it wants in your web browser defeats the point. CORS does not help. Also, the firewall has to block most DNS traffic, otherwise the model could query `A <secret>.evil.com` and Google/Cloudflare servers (along with everybody else) will forward the query to evil.com. Secure DNS, therefore, also can't be allowed. katakate[1] is still incomplete, but something that it is the solution here. Run the LLM and its code in firewalled VMs. [1]: https://github.com/Katakate/k7

As promised: https://github.com/Katakate/k7/tree/fix/no-dns-res-in-lockdo... Will merge that in after it passes all network tests on a clean/wiped instance.

Lol.. a full-fledged project just launched on Github using firecracker and giving cli/api interfaces: https://github.com/Katakate/k7

I've built this to make it easy to host your own infra for lightweight VMs at large scale. Intended for exec of AI-generated code, for CICD runners, or for off-chain AI DApps. Mainly to avoid Docker-in-Docker dangers and mess. Super easy to use with CLI / Python SDK, friendly to AI engs who usually don't like to mess with VM orchestration and networking too much. Defense-in-depth philosophy. Would love to get feedback (and contributors: clear & exciting roadmap!), thx

I built Katakate to make it easy for AI engineers to run workloads in lightweight, isolated VMs. It’s useful for agentic runs, evals, or any setup where you want safe sandboxing at scale. - Dozens of VMs per node with low overhead - Python SDK included - Docs: https://docs.katakate.org - LangChain agent tutorial: https://docs.katakate.org/guides/langchain-agent - Deepwiki explainer: https://deepwiki.com/katakate/k7 I’d appreciate feedback on developer experience and whether this fits into your AI workflows. Tech stack: Kubernetes, Kata, Firecracker.