Infrastructure & Builders | Friday
Cloudflare ships accountless agent deployment — any agent can now provision live infrastructure in seconds
2026-06-19 · Source
Cloudflare released Temporary Accounts for Workers on June 19, enabling AI agents to deploy live serverless infrastructure without an account, OAuth flow, or human credential setup. A single CLI command — wrangler deploy --temporary — provisions a real Cloudflare account, issues an API token, deploys the Worker, and returns a claim URL. The deployment stays live for 60 minutes. A human can claim it within that window to make it permanent; unclaimed deployments and all bound resources are automatically deleted. Supported resources at launch include Workers, KV, D1, Durable Objects, Hyperdrive, and Queues. The Wrangler CLI was updated to surface the flag automatically when an agent attempts a deploy without credentials.
Why it matters: Agents can now complete the full write-deploy-verify loop on live infrastructure without a human unlocking the environment first — removing the last hard stop between autonomous code generation and production deployment.
Brief Items
Mastercard launched Agent Pay for Machines with 31 partners including Stripe, Coinbase, and Cloudflare
Mastercard launched Agent Pay for Machines (AP4M) on June 10, a live protocol enabling AI agents to transact autonomously across Mastercard's global payment rails. The system covers four functions: credentialing registered agents, defining spend permissions, transacting across card and account infrastructure, and settling in fiat or stablecoins. Thirty-one launch partners include Stripe, Coinbase, Adyen, Cloudflare, BVNK, Checkout.com, and OKX. Read more
Cloudflare added GLM-5.2 agentic reasoning model to Workers AI with 262K token context
Cloudflare integrated the GLM-5.2 model into Workers AI, featuring a 262,144 token context window and native function calling support. The model is optimized for reasoning and coding tasks required by autonomous agents. It runs on Cloudflare's global edge network for low-latency inference without external API calls. Read more
Agent Commerce Weekly · agentcommerceweekly.com