AI BriefingCursor Cloud Agents Support Vercel Sandbox Execution Environment
Key point
Cursor Cloud Agents now support execution on Vercel Sandbox's Firecracker microVMs, providing a scalable execution environment for Enterprise plan users.
Details
Cursor Cloud Agents can now run on Vercel Sandbox instead of the existing Cursor-hosted machines. Cursor manages the agent harness and inference loop, and supports users provisioning their own execution environments via the Self-Hosted Machines API. This feature is available to Cursor Enterprise plan users.
Vercel Sandbox provides an isolated Firecracker microVM as the execution environment for each agent request. Vercel Functions and Vercel Workflow form a durable control plane to handle pending agent requests, provision workers, monitor sessions, and perform automatic cleanup.
This architecture offers the following benefits:
- Operating scale-to-zero worker pools without long-running virtual machines
- Assigning a dedicated and isolated Sandbox for every agent request
- Durable retries upon worker or session failures
- Managing short-lived, user-scoped credentials within each Sandbox
Users can follow the step-by-step guide to deploy the reference implementation to their own Vercel account, or check the Vercel Sandbox documentation for detailed information.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Perplexity: AI Agents Bypass SPACE Network Policies; Direct Tests Reveal Weaknesses in 7 of 9 Third-Party Sandboxes
AWS Introduces Enhanced Custom Event Buses for Organization-Wide Event Sharing