--- title: Introducing Experimental Workflows and Orchestrators in... description: TanStack has released an experimental PR build (not merged to main) called TanStack AI Workflows & Orchestrators, available via pkg.pr.new. It lets developers... canonical: https://daily.dev/posts/introducing-experimental-workflows-and-orchestrators-in-tanstack-ai-yzurhsyy3 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Introducing Experimental Workflows and Orchestrators in TanStack AI | daily.dev og:description: TanStack has released an experimental PR build (not merged to main) called TanStack AI Workflows & Orchestrators, available via pkg.pr.new. It lets developers... og:url: https://daily.dev/posts/introducing-experimental-workflows-and-orchestrators-in-tanstack-ai-yzurhsyy3 og:image: https://api.daily.dev/og/posts/yzuRHsYy3.png og:image:alt: Introducing Experimental Workflows and Orchestrators in TanStack AI og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Introducing Experimental Workflows and Orchestrators in TanStack AI **[TanStack](https://daily.dev/sources/tanstack-blog)** · 11 min read · 0 upvotes · 0 comments ## Summary TanStack has released an experimental PR build (not merged to main) called TanStack AI Workflows & Orchestrators, available via pkg.pr.new. It lets developers compose typed LLM/agent steps as TypeScript async generators using defineAgent and defineWorkflow, with support for streaming to the UI via SSE, pausing for human approval, and resuming runs. New packages include @tanstack/ai-orchestration with runWorkflow, parseWorkflowRequest, and inMemoryRunStore, plus React hooks useWorkflow and useOrchestration. Orchestrators use a router (defineOrchestrator, defineRouter) to pick the next agent turn by turn instead of following a fixed sequence. The API is explicitly unstable and open for community feedback via GitHub PR 542 before stabilization. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How do I install the experimental TanStack AI orchestration package before it's officially released? Install it directly from the PR build hosted on pkg.pr.new rather than npm, since it has not been merged to main. Run commands like `npm i https://pkg.pr.new/TanStack/ai/@tanstack/ai-orchestration@542` along with matching PR versions of @tanstack/ai, @tanstack/ai-react, @tanstack/ai-client, and @tanstack/ai-openai, all tied to PR 542. _daily.dev surfaces early framework previews like this for teams evaluating new AI tooling before it stabilizes._ ### How do I pause an AI workflow for human approval and resume it later with TanStack AI orchestration? Call `yield* approve({ title, description })` inside a workflow generator; the runtime emits an approval-requested event, persists the paused run in the run store, and closes the SSE connection without keeping a socket open. When the client later calls approve() with a result, it POSTs back to the same endpoint with the run id, and runWorkflow resumes the generator from the paused point. _developers building approval gates into AI pipelines can track evolving patterns like this on daily.dev._ ### What's the difference between a workflow and an orchestrator in TanStack AI? A workflow defines a fixed sequence of typed agent steps directly in an async generator body using yield*, best when the pipeline order is known in advance, such as writer then editor then approval. An orchestrator instead uses a router function (defineRouter) that inspects input, state, and the previous result each turn to decide which agent runs next, suited for cases where the next step isn't predetermined. _daily.dev helps developers comparing agent orchestration patterns like workflows versus routers stay current._ ## Similar posts on daily.dev - [TanStack Releases Framework Agnostic AI Toolkit](https://daily.dev/posts/tanstack-releases-framework-agnostic-ai-toolkit-j747yitl6) · InfoQ · 1 upvotes · 0 comments - [Introducing Agent Kit: Storage Workflows for AI Agents](https://daily.dev/posts/introducing-agent-kit-storage-workflows-for-ai-agents-537zvqz61) · Tigris · 1 upvotes · 1 comments - [My Experience with GitHub Agentic Workflows](https://daily.dev/posts/my-experience-with-github-agentic-workflows-orst3dlvf) · Thomas Thornton · 0 upvotes · 0 comments --- Tags: [#react](https://daily.dev/tags/react), [#ai-agents](https://daily.dev/tags/ai-agents), [#typescript](https://daily.dev/tags/typescript), [#tanstack](https://daily.dev/tags/tanstack) [View this post on daily.dev](https://daily.dev/posts/introducing-experimental-workflows-and-orchestrators-in-tanstack-ai-yzurhsyy3) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Introducing Experimental Workflows and Orchestrators in TanStack AI","url":"https://daily.dev/posts/introducing-experimental-workflows-and-orchestrators-in-tanstack-ai-yzurhsyy3","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/introducing-experimental-workflows-and-orchestrators-in-tanstack-ai-yzurhsyy3"},"datePublished":"2026-08-23T12:23:56.452Z","dateModified":"2026-08-23T12:57:08.020Z","description":"TanStack has released an experimental PR build (not merged to main) called TanStack AI Workflows & Orchestrators, available via pkg.pr.new. It lets developers...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/e149cf66271c4adfc2552dbf33ad7334?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/e149cf66271c4adfc2552dbf33ad7334?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"TanStack","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"TanStack","logo":"https://media.daily.dev/image/upload/s--TVIrvIT7--/f_auto,q_auto/v1787487634/logos/tanstack-blog?_a=BAMAMicg0","url":"https://daily.dev/sources/tanstack-blog"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/introducing-experimental-workflows-and-orchestrators-in-tanstack-ai-yzurhsyy3","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"react,ai-agents,typescript,tanstack","timeRequired":"PT11M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"TanStack","item":"https://daily.dev/sources/tanstack-blog"},{"@type":"ListItem","position":3,"name":"Introducing Experimental Workflows and Orchestrators in TanStack AI"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/introducing-experimental-workflows-and-orchestrators-in-tanstack-ai-yzurhsyy3#faq","mainEntity":[{"@type":"Question","name":"How do I install the experimental TanStack AI orchestration package before it's officially released?","acceptedAnswer":{"@type":"Answer","text":"Install it directly from the PR build hosted on pkg.pr.new rather than npm, since it has not been merged to main. Run commands like `npm i https://pkg.pr.new/TanStack/ai/@tanstack/ai-orchestration@542` along with matching PR versions of @tanstack/ai, @tanstack/ai-react, @tanstack/ai-client, and @tanstack/ai-openai, all tied to PR 542. daily.dev surfaces early framework previews like this for teams evaluating new AI tooling before it stabilizes."}},{"@type":"Question","name":"How do I pause an AI workflow for human approval and resume it later with TanStack AI orchestration?","acceptedAnswer":{"@type":"Answer","text":"Call `yield* approve({ title, description })` inside a workflow generator; the runtime emits an approval-requested event, persists the paused run in the run store, and closes the SSE connection without keeping a socket open. When the client later calls approve() with a result, it POSTs back to the same endpoint with the run id, and runWorkflow resumes the generator from the paused point. developers building approval gates into AI pipelines can track evolving patterns like this on daily.dev."}},{"@type":"Question","name":"What's the difference between a workflow and an orchestrator in TanStack AI?","acceptedAnswer":{"@type":"Answer","text":"A workflow defines a fixed sequence of typed agent steps directly in an async generator body using yield*, best when the pipeline order is known in advance, such as writer then editor then approval. An orchestrator instead uses a router function (defineRouter) that inspects input, state, and the previous result each turn to decide which agent runs next, suited for cases where the next step isn't predetermined. daily.dev helps developers comparing agent orchestration patterns like workflows versus routers stay current."}}]} ```