What problem does it solve? Building AI agents that persist state, handle real-time WebSocket connections, run scheduled tasks, and survive restarts requires complex infrastructure. This Skill provides up-to-date guidance for the Cloudflare Agents SDK, prioritizing live documentation retrieval over potentially outdated pre-trained knowledge. ## Core Features & Use Cases - Stateful Agents: Create Agent classes with SQLite-backed state, @callable RPC methods, scheduling, queues, retries, and durable execution via runFiber. - Chat & Streaming: Build streaming chat agents with AIChatAgent, resumable streams, tool calling, human-in-the-loop approvals, and React hooks (useAgent, useAgentChat). - Integrations: Connect to or build MCP servers, handle email routing, webhooks, push notifications, Workflows for long-running tasks, plus experimental voice, browser automation, and Think agents. - Use Case: Build a customer support chat agent that persists conversation history, schedules follow-up reminders, calls MCP tools, and streams responses to a React frontend — all deployed as a Cloudflare Worker with Durable Objects. ## Quick Start Ask the AI to create a Cloudflare Workers agent using the Agents SDK with a Counter class that has callable increment and decrement methods.