Cloudflare, Anthropic Launch Secure AI Agent Platform
Cloudflare and Anthropic have partnered to launch Cloudflare Environments for Claude Managed Agents. The collaboration allows organizations to run autonomous AI workloads on Cloudflare’s global infrastructure while continuing to manage agent orchestration through Anthropic’s Claude platform.
The integration extends support for Claude Managed Agents, which entered public beta earlier this year. As a result, developers can now execute workloads inside Cloudflare environments instead of depending entirely on Anthropic-hosted systems.
Flexible Agent Execution and Custom Tools
The platform includes a default deployment template with several infrastructure options for AI agents. Developers can run workloads inside lightweight V8 isolates for faster startup times and lower operating costs. Alternatively, they can use Linux microVMs when workloads require full operating system capabilities.
Additionally, the system supports secure browser automation, session recording, and custom tool integration with minimal configuration. Developers can also add utilities such as git and Bash when container-based environments become necessary.
Cloudflare integrated private connectivity through Zero Trust proxies. Consequently, organizations can securely connect AI agents to internal services without exposing infrastructure to the public internet.
Security and Monitoring Features
Security remains a central focus of the integration. Cloudflare’s outbound proxy architecture allows organizations to define network access policies for specific agents or tenants. Furthermore, the platform supports endpoint allowlists and encrypted credential injection without directly exposing secrets to AI agents.
The environment also includes detailed observability tools. Organizations can review browser session records, monitor sandbox performance metrics, and analyze execution logs in real time. In addition, developers can access running environments through SSH for debugging and operational troubleshooting.
The announcement builds on Cloudflare’s broader expansion into AI infrastructure services. Earlier this year, the company brought its Sandboxes and Containers products into general availability with features focused on secure credential management, terminal access, and compute-based pricing models designed for large-scale AI agent deployment.
