Hi everyone,
We are excited to announce the release of Agent Sandbox v0.5.0! This major milestone marks the official graduation of our core and extension APIs to v1beta1, delivering critical security hardening, enhanced platform stability, and several new features across the platform, SDKs, and examples.
Key highlights:
- API Graduation and Deprecation (
v1alpha1 to v1beta1): The core and extension APIs (agents.x-k8s.io and extensions.agents.x-k8s.io) have graduated to v1beta1. The v1alpha1 APIs are now officially deprecated.Please update your manifests to target the new versions. For detailed step-by-step upgrade instructions, please refer to the API Migration Guide. - Security Hardening: Disabled automatic HTTP redirects in the Python and Go SDKs to defend against Server-Side Request Forgery (SSRF) attacks.We also resolved an unauthenticated internal proxy vulnerability in the sandbox router and blocked IPv6 link-local traffic by default.
- Performance and Warm Pools: Implemented optimized, parallel scale operations within the
SandboxWarmPool controller, significantly speeding up Sandbox creation and deletion. - Client SDK Improvements: SDKs now support dynamic timeout propagation to the sandbox router, and the Go client has enabled PodIP routing.
- New Examples & Documentation: Added examples for multi-cluster warm-pool orchestration of Reinforcement Learning (RL) and Evals workloads, running Anthropic Managed Agents, and running an MCP server inside a sandbox with persistent storage.
We would love to get your feedback and welcome you to get involved:
- Slack: Join the discussion in the #agent-sandbox channel on Kubernetes Slack.
- Meetings: We host our community meetings biweekly on Mondays at 9:00 PT (Pacific Time). You can add it to your schedule using our Meeting Calendar.
You can find more info on our GitHub Repository and our Docs Site.
A huge thank you to all the contributors who helped make this release possible!