openai-agents

Build AI applications with text agents, voice agents (realtime), multi-agent workflows, tools, guardrails, and human-in-the-loop patterns. Breaking Change (v0.4.0) : Zod 3 no longer supported. Upgrade to zod@4 . Runtimes : Node.js 22+, Deno, Bun, Cloudflare Workers (experimental) Agents : LLMs with instructions + tools Tools : Functions with Zod schemas Handoffs : Multi-agent delegation Guardrails : Input/output validation Structured Outputs : Type-safe responses Basic : const result = await run(agent, 'What is 2+2?') Streaming : Agent-as-Tool Context Isolation : When using agent.asTool() , sub-agents do NOT share parent conversation history (intentional design to simplify debugging). Workaround : Pass context via tool parameters: Source : Issue #806 Input : Validate before processing Output : Filter responses (PII detection, content safety) Streaming HITL : When using stream: true with requiresApproval , must explicitly check interruptions: Example : human-in-the-loop-stream.ts Create : Browser Session : CRITICAL : Never send OPENAI_API_KEY to browser! Generate ephemeral session tokens server-side. Voice Handoffs : Voice/model must match across agents (cannot change during handoff) Limitations : Templates : References : Cloudflare Workers (experimental): Limitations : Next.js : app/api/agent/route.ts → POST handler with run(agent, message) Templates : cloudflare-workers/ ,...