AI BriefingVercel Sandbox Adds Support for Forking
Key point
Vercel Sandbox now supports a forking feature that creates a new instance by replicating the snapshot and configuration of an existing sandbox.
Details
Vercel Sandbox has started supporting Forking functionality through the Sandbox.fork() method. Forking lets you create a new sandbox starting from the source's current snapshot, inheriting its config and environment variables as-is.
If the source is running, forking is based on the latest saved state; if there is no snapshot, a new sandbox is created using the source's runtime and settings. You can freely override any inherited values via the parameters you pass in.
Key use cases include:
- Branching AI agents from a shared base
- Providing each tenant with an individual copy of a template
- Running multiple variants in parallel based on a single configuration
Users can easily perform forking by calling Sandbox.fork() in the SDK, or by using the sandbox fork command in the CLI.
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
Worktrunk CLI Released for Managing Git Worktrees for Parallel AI Coding AgentsDevin Adds Cloud Session Control and SSH Access from Terminal