From aa601d3d56f760e4ba9ae83787d5e42a69ccfd69 Mon Sep 17 00:00:00 2001 From: Tanner Linsley Date: Sat, 22 Aug 2026 10:23:57 -0600 Subject: [PATCH 1/2] Clarify TanStack AI durability setup --- src/blog/tanstack-ai-rc.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/blog/tanstack-ai-rc.md b/src/blog/tanstack-ai-rc.md index 2585b2bc9..822699dcb 100644 --- a/src/blog/tanstack-ai-rc.md +++ b/src/blog/tanstack-ai-rc.md @@ -31,7 +31,7 @@ Our initial goal was to create an architecture that could be easily extended not One of the biggest payoffs since we started has been making the right architectural decisions early and building everything on top of those foundations. One of the strongest parts of that architecture is our [middleware system](https://tanstack.com/ai/latest/docs/advanced/middleware). -You can build just about anything with middleware. Chat persistence, agent harnesses running inside sandboxes, memory, telemetry, durability, and many other features are simply middleware functions that you pass into your `chat()` method. +You can build just about anything with middleware. Chat persistence, agent harnesses running inside sandboxes, memory, telemetry, and many other features are middleware functions that you pass into your `chat()` method. Durability is separate and plugs into the streaming response through an adapter. Another powerful feature we offer is [lazy tool calling](https://tanstack.com/ai/latest/docs/tools/lazy-tool-discovery), which can help reduce token costs with a simple Boolean flag. We also support advanced [interrupt scenarios](https://tanstack.com/ai/latest/docs/interrupts/overview) driven by your schemas. You can modify tool-call arguments, trigger multiple interrupts, and mix and match these capabilities while keeping everything completely type-safe from end to end. From 0745211a45a2760ec68ff288a048c1f2f8f196b3 Mon Sep 17 00:00:00 2001 From: Tanner Linsley Date: Sat, 22 Aug 2026 10:29:43 -0600 Subject: [PATCH 2/2] Use accurate AI middleware terminology --- src/blog/tanstack-ai-rc.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/blog/tanstack-ai-rc.md b/src/blog/tanstack-ai-rc.md index 822699dcb..21efe7d74 100644 --- a/src/blog/tanstack-ai-rc.md +++ b/src/blog/tanstack-ai-rc.md @@ -31,7 +31,7 @@ Our initial goal was to create an architecture that could be easily extended not One of the biggest payoffs since we started has been making the right architectural decisions early and building everything on top of those foundations. One of the strongest parts of that architecture is our [middleware system](https://tanstack.com/ai/latest/docs/advanced/middleware). -You can build just about anything with middleware. Chat persistence, agent harnesses running inside sandboxes, memory, telemetry, and many other features are middleware functions that you pass into your `chat()` method. Durability is separate and plugs into the streaming response through an adapter. +You can build just about anything with middleware. Chat persistence, agent harnesses running inside sandboxes, memory, telemetry, and many other features are middleware you pass into your `chat()` method. Durability is separate and plugs into the streaming response through an adapter. Another powerful feature we offer is [lazy tool calling](https://tanstack.com/ai/latest/docs/tools/lazy-tool-discovery), which can help reduce token costs with a simple Boolean flag. We also support advanced [interrupt scenarios](https://tanstack.com/ai/latest/docs/interrupts/overview) driven by your schemas. You can modify tool-call arguments, trigger multiple interrupts, and mix and match these capabilities while keeping everything completely type-safe from end to end.