Simon Willison publishes external sandbox approach for coding agents
What happened
On 18 November 2026, Simon Willison, working with AI research lab Prime Radiant, proposed an "external sandbox" model for coding agents: the code an agent executes runs inside a separate sandbox, while the agent itself stays outside it. The report does not disclose implementation details, scope or test results, nor whether an earlier version or pilot existed. The confirmed development is the publication of the approach; deployment and evaluation await further information.
Written by AI from the reports · updated 19 h ago
Timeline
Follow the reports to see the event from each side.
- Simon WillisonExternal sandboxes for coding agents
External sandboxes for coding agents I've been exploring new ways to sandbox code that is executed by agents, with a focus on making it safer to try out local models without worrying that a mistake or prompt injection attack might cause them to take damaging actions on a machine.
Heat of this event
Current heat 5·Peak in the comparable range 9(Nov 18 11:00)·Change over 24 hours in the comparable range –
The trend compares only the same accounts observed without a break, so its range can be narrower than the current heat count. Move the pointer or tap the chart to see each hour; with a keyboard, use the left and right arrow keys.