exec-sandbox released for running untrusted agent code in QE
Start here
exec-sandbox has been released as a self-hosted Python asyncio library and sbx CLI that runs untrusted Python, JavaScript or shell code — such as code written by an AI agent — in QEMU microVMs on the user's own Linux or macOS hosts. It requires no cloud account or API key, and the first use downloads the VM images from GitHub Releases. The project claims starts under 1 ms from a warm pool and about 50–60 ms from an L1 memory snapshot. Its save() and reuse() calls store encrypted VM state in S3 or a local folder, so instances can resume later on any worker of the same host kind. Network access is off by default. On Windows, users must run it under WSL2; importing exec_sandbox on native Windows raises ConfigError PLATFORM_UNSUPPORTED.
Generated from reports · updated 3 h ago
Report timeline
Follow the reports to see the event from different sides.
- Hacker News · AIexec-sandbox launches a self-hosted QEMU microVM sandbox for AI agent code
exec-sandbox is a Python asyncio library and sbx CLI that runs untrusted Python, JavaScript or shell code in QEMU microVMs on local Linux or macOS hosts, with no cloud account or API key required. It claims under 1 ms starts from a warm pool and about 50-60 ms from an L1 memory snapshot, and its save() and reuse() calls store encrypted VM state in S3 or a local folder so instances can resume later on any worker of the same host kind.
Heat for this event
- Current, comparable range
- 9
- Peak, comparable range
- 10November 18 05:00
- Change in 24 hours
- –
The trend compares the same participants over hours observed continuously, so its range can be narrower than the current heat count. Hover or tap the chart for hourly heat; the arrow keys move through it.