WDK Resumable Streaming
AI chat UIs and agent dashboards need two things at once: streaming responses for perceived speed, and durable state so a reload doesn't lose the conversation. Workflow DevKit's resumable streams give you both — the same writer that powers `streamText`-style UX is also the workflow's persisted output. If you've ever lost an in-progress LLM response to a refresh, this is the fix.
What It Defines
Mechanism for delivering data from a long-running workflow to a client (browser, mobile app, agent client) before the workflow completes — and resuming the stream cleanly across reconnects. Backed by the event log: every chunk written to the stream is persisted, so a reconnecting client picks up exactly where it left off rather than restarting the workflow or losing tokens. Designed for AI streaming (LLM token deltas, tool-call traces) and for any "render partial progress as it happens" UI on top of a durable backend.
Canonical (Normative)
Frontend cloud platform and the steward of the open-source Workflow DevKit (Apache-2.0). Vercel publishes the WDK spec and reference implementation ("use workflow" / "use step" directives, durable timers, hooks, streaming) and runs the integrated workflow runtime on its platform. WDK can also be self-hosted on Docker, AWS, or DigitalOcean. Vercel also publishes the AI SDK and the Vercel Functions runtime spec.
Related Specs
WDK is the durable-execution model arriving inside the JavaScript ecosystem proper: instead of writing a state machine, you write `async` code with two extra string directives and you get crash-safe, resumable, retry-aware, observable workflows. If you're building AI agents, multi-step background jobs, human-in-the-loop flows, or anything that previously needed BullMQ + a state machine + careful idempotency, WDK collapses that into a single mental model. Even if you don't pick WDK specifically, the directives + replay model are the durable-execution pattern Temporal pioneered and Cloudflare/Restate/DBOS now ship — knowing one teaches you the rest.
If you remember one thing about durable execution, remember this: workflow code runs many times and the event log decides what's real. That single fact explains every WDK rule — why steps must be idempotent, why you can't read `Date.now()` directly in a workflow, why parameters are passed by value, why mutating an object inside a step doesn't propagate, and why bundlers occasionally bite you. Internalize the replay model and the rest of WDK (and Temporal, and Cloudflare Workflows) becomes obvious.
The go-to protocol for server-push without WebSocket complexity. Used for AI/LLM streaming responses, live notifications, and real-time dashboards over plain HTTP.
These are the primitives that make durable workflows worth their cost. Anything you'd previously implement with a cron job, a state column, a polling loop, or a queue + scheduled job collapses into ordinary linear code. Human-in-the-loop, long-running approval flows, multi-day onboarding sequences, and "call a tool then wait for the user" agent patterns all fall out of `sleep` + `createWebhook` for free.