OpenAI Agents SDK
Ai ToolOpenAI Agents SDK is a developer software-development kit for building, testing and instrumenting AI agents that can run on OpenAI and other LLM back-ends. It provides runtime hooks, tracing and tool-call interfaces so developers can integrate multi-model agents into applications such as ChatGPT Apps or external runtimes.
Stories
Completed digest stories linked to this service.
-
Copilot SDK adds session memory and lazy tool loading2026-06-18GitHub Copilot SDK now supports persistent session memory and lazy tool loading that leans on model tool searc...
-
OpenAI bakes in observability for agents: Codex 0.135.0 + Agents JS 0.11.62026-05-29OpenAI tightened observability and state handling for agent workflows with Codex 0.135.0 and openai‑agents‑js ...
-
MCP agents get safer: OpenAI Agents SDK 0.10.1 validates policies, fixes history...2026-05-08OpenAI Agents SDK 0.10.1 tightens MCP agent safety with approval-policy validation and fixes session history l...
-
Codex 0.122 hardens sandboxing and plugins, smooths TUI; watch for upgrade snags2026-04-21OpenAI’s Codex app 0.122 ships stronger sandboxing, revamped plugin workflows, and TUI upgrades, with a few up...
-
Sandboxed coding agents: OpenAI updates its Agents SDK, and there’s a clear way ...2026-04-19OpenAI’s Agents SDK now includes sandboxing and a model harness, and there’s a practical way to benchmark agen...
-
Plan for multi-model agents and resilience in 20262026-02-03AI agents are set to pressure reliability, with more outages expected and a push toward chaos engineering and ...
Resources
Links to check for updates: homepage, feed, or git repo.