Scale AI's AgentEnv Now Runs on Vercel Sandbox
Article Content
- •AgentEnv integration with Vercel Sandbox enables isolated testing environments.
- •The demo showcases a deterministic agent performing tasks without language model calls.
- •Users need specific software versions and configurations to run evaluations.
Scale AI's AgentEnv framework has been integrated with Vercel Sandbox, allowing users to run evaluations of agents against tasks in isolated environments. The integration is available in AgentEnv version 0.9.1334 and later. Users can deploy a complete evaluation using Vercel's infrastructure, which includes separate sandboxes for agents and their task environments. The demo provided showcases a deterministic agent performing tasks without calling a language model, focusing on validating the integration rather than model quality. The setup requires Git, Python 3.11.4 or newer, Node.js, and Vercel CLI 62.5.0 or newer. This integration is aimed at developers looking to leverage Vercel's capabilities for testing AI agents effectively.
Ask AI about this cluster
Answers cite the sources they use
Timeline
More articles in this cluster (3)
Following this threat?
Track Vercel in your own feed — you're alerted when they show up in new reporting, leak sites or exploitation.
Free account · no card needed
Common questions
What is AgentEnv?
What are the prerequisites for using AgentEnv on Vercel?
How does the demo validate the integration?
Continue Reading
Critical RCE Vulnerability in Confection WordPress Plugin (CVE-2026-115252) A critical vulnerability, CVE-2026-115252, has been identified in the Confection WordPress Cable plugin, affecting versions 0.4.0 to 0.4.7. This flaw allows unauthenticated attackers to execute arbitrary PHP code on the server by writing unmodified settings into a PHP file without proper checks. The vulnerability has…