32 articles
Cloudflare launches the Registrar API in beta, enabling programmatic domain search, availability checking, and registration.
Cloudflare has launched Browser Run, enabling AI agents to interact with the web at scale.
Cloudflare releases an experimental voice pipeline for the Agents SDK that enables real-time voice interaction with agents.
Cloudflare's Managed OAuth enables agents to securely access internal applications protected by Cloudflare Access with a single click.
This post discusses securing non-human identities like AI agents and scripts through automated token revocation, OAuth management, and resource-scoped permissions.
Cloudflare Sandboxes is now generally available, providing AI agents with isolated environments to develop and run code safely at scale.
Outbound Workers for Sandboxes enable secure, identity-aware authentication for sandboxed agents without exposing credentials to untrusted workloads.
Cloudflare examines how AI agents are fundamentally changing cloud infrastructure requirements from traditional one-to-many to one-to-one models.
This post introduces Cloudflare's Dynamic Worker Loader, an open beta API that sandboxes AI-generated code using V8 isolates instead of containers, enabling consumer-scale agent execution.
Cloudflare's Workers AI now supports large frontier open-source models, launching with Moonshot AI's Kimi K2.5 to power agentic workloads on a unified platform.
Cloudflare introduces Markdown for Agents, a feature that automatically converts HTML pages to markdown at the network level for AI crawlers and agents.
This post introduces Moltworker, a middleware solution that enables running Moltbot (a self-hosted personal AI agent) on Cloudflare's infrastructure without dedicated hardware.
