NOFire Open-Sources Brig, an Apache 2.0 MicroVM Sandbox for AI Coding Agents
On September 15, 2026, NOFire AI released Brig under Apache 2.0 to run AI coding agents inside hardware-enforced microVMs on macOS and Linux, with install via brig.sh and curated profiles for major agent CLIs.
NOFire AI open-sourced Brig on September 15, 2026 as an Apache 2.0 project that runs AI coding agents inside a low-overhead microVM on macOS and Linux. The company’s press release frames the gap as agents that can install packages, hit the network, and touch developer credentials when teams turn on auto-approval—risks Brig aims to contain inside an ephemeral, hardware-enforced virtual machine boundary.
Install is documented as curl -fsSL https://brig.sh/install | sh. The release says Brig ships curated profiles for Claude Code, Codex, Cursor, Gemini, Grok, and opencode, and can also boot bring-your-own OCI images such as Ubuntu. Documentation and the security model are pointed at brig.sh and the project repositories.
License and audit surface
NOFire says all Brig components are Apache 2.0 and that the microVMM code a security team would need to audit is under 20,000 lines. The Go module listing for github.com/brig-sh/brig also reports Apache-2.0. On macOS, Brig drives the companion hull microVM runtime (Virtualization.framework); github.com/brig-sh/hull is likewise listed under Apache License 2.0. On Linux, public package docs describe a nerdctl/containerd path with the urunc shim so the sandbox remains a microVM rather than a plain container by default.
This brief is a product/open-source launch note, not a vulnerability write-up—no reproduction steps for agent escapes or supply-chain attacks.
Primary sources are NOFire’s September 15 press release, brig.sh install guidance, and the Apache-2.0 listings for brig-sh/brig and brig-sh/hull.
- #Cyber Security
- #AI Agents
- #Opensource
Author
Raj M
Contributor
AI Systems Architect is a seasoned technology leader with over 15 years of experience in the IT industry working with Fortune 500 companies. With a solid foundation in multi-agent systems, open-source LLM infrastructure, and enterprise deployment, he excels at building scalable production-grade AI platforms.