diff --git a/.changeset/ag-ui-interrupts.md b/.changeset/ag-ui-interrupts.md
deleted file mode 100644
index a3a23ff85..000000000
--- a/.changeset/ag-ui-interrupts.md
+++ /dev/null
@@ -1,21 +0,0 @@
----
-'@tanstack/ai': minor
-'@tanstack/ai-client': minor
-'@tanstack/ai-react': minor
-'@tanstack/ai-preact': minor
-'@tanstack/ai-solid': minor
-'@tanstack/ai-vue': minor
-'@tanstack/ai-svelte': minor
-'@tanstack/ai-angular': minor
----
-
-Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
-client-tool execution, with typed bound resolvers, atomic batches, and
-structured errors. Interrupts run ephemerally by resuming from the full client
-message history in a fresh child run — no persistence required.
-
-This changes native approval and client-tool streams from legacy custom events
-to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
-`pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
-legacy event readers remain as limited compatibility surfaces for migration;
-`addToolResult` remains supported.
diff --git a/.changeset/devtools-memory-inspector.md b/.changeset/devtools-memory-inspector.md
deleted file mode 100644
index 5fd1e33aa..000000000
--- a/.changeset/devtools-memory-inspector.md
+++ /dev/null
@@ -1,29 +0,0 @@
----
-'@tanstack/ai-memory': minor
-'@tanstack/ai-event-client': minor
-'@tanstack/ai-client': minor
-'@tanstack/ai-devtools-core': minor
----
-
-**Surface server-side memory state in the TanStack AI DevTools.**
-
-The DevTools panel now has a **Memory** tab for any chat wired with
-`memoryMiddleware`. It shows, per scope (session), an operations timeline (each
-turn's recall — query, fragment count, injected system-prompt size, whether
-memory tools were exposed, duration) and the current stored records/facts when
-the adapter implements the optional `inspect`/`listFacts` methods.
-
-Because memory runs on the server (whose event bus never reaches the browser),
-the middleware transports its state to the panel over the chat stream as a
-`memory:state` `CUSTOM` event, which `@tanstack/ai-client`'s devtools bridge
-re-emits as browser `memory:*` events — the same pattern generation results use.
-The snapshot reflects memory as of the start of each turn; opening the panel
-mid-conversation replays the latest state so the tab isn't empty.
-
-- `@tanstack/ai-memory` — `memoryMiddleware` injects a `memory:state` `CUSTOM`
-  chunk carrying recall metrics + an `inspect`/`listFacts` snapshot; exports
-  `MEMORY_STATE_EVENT` and `MemoryStateEventValue`.
-- `@tanstack/ai-event-client` — adds the `memory:snapshot` devtools event.
-- `@tanstack/ai-client` — the chat devtools bridge re-emits `memory:*` from the
-  transported chunk and replays the last snapshot on `devtools:request-state`.
-- `@tanstack/ai-devtools-core` — new Memory tab + per-scope memory store slice.
diff --git a/.changeset/interrupt-binding-ownership.md b/.changeset/interrupt-binding-ownership.md
deleted file mode 100644
index b681fe9dd..000000000
--- a/.changeset/interrupt-binding-ownership.md
+++ /dev/null
@@ -1,33 +0,0 @@
----
-'@tanstack/ai-client': minor
-'@tanstack/ai': minor
----
-
-Make interrupt ownership explicit rather than assumed.
-
-An AG-UI `Interrupt` is a shared envelope — a workflow engine's durable
-approval or another agent framework's pause can arrive on the same stream. What
-makes a pause resumable through `chat()` is the binding this package attaches
-under `tanstack:interruptBinding`.
-
-- Interrupts that carry no binding this client understands now surface as
-  `kind: 'unbound'` with `canResolve: false`, instead of being given a
-  synthesized binding and rendered as resolvable generic interrupts. Resolving
-  those produced an answer submitted against a run with nothing pending, which
-  failed as `unknown-interrupt` only after the user had filled in the form.
-  Unbound items never block submission of the interrupts that are yours.
-- The binding carries a wire version (`INTERRUPT_BINDING_VERSION`). Readers
-  reject a version they don't recognise rather than duck-typing its fields. A
-  binding written before the field existed is still read.
-- `INTERRUPT_BINDING_METADATA_KEY`, `withInterruptBinding()` and
-  `readInterruptBinding()` are exported, so anything producing an interrupt this
-  package must later resume attaches the binding through a supported API
-  instead of copying the metadata key.
-- Interrupt classification is driven by the binding alone. `Interrupt.reason` is
-  free-form AG-UI text another producer can also use, so it is now a display
-  hint only and never decides ownership.
-- The interrupt protocol surface is enumerated instead of `export *`. The
-  unimplemented durable-recovery contract (`InterruptRecoveryStateV1`,
-  `InterruptRecoveryQuery`, the never-called `loadInterruptState` adapter hook,
-  and the `persistence-required` / `atomic-commit-unsupported` /
-  `recovery-unavailable` error codes) is removed rather than published.
diff --git a/.changeset/interrupts-validation-ownership.md b/.changeset/interrupts-validation-ownership.md
deleted file mode 100644
index 122b0687b..000000000
--- a/.changeset/interrupts-validation-ownership.md
+++ /dev/null
@@ -1,20 +0,0 @@
----
-'@tanstack/ai': minor
-'@tanstack/ai-client': minor
----
-
-Interrupts: the application owns wire-schema validation, and the hashing
-dependency is gone.
-
-The library no longer transforms a generic interrupt's wire JSON Schema into a
-validator or validates the resolved value against it, on either the client or
-the server. Whatever you pass to `resolveInterrupt` (client) or send in the
-`resume` batch (server) flows through as-is. Validate it yourself if you need to
-trust it, e.g. with `z.fromJSONSchema(interrupt.responseSchema).safeParse(value)`
-on the client and your own check on the server. Validation of a tool's
-code-authored Standard Schema (`approvalSchema` / `inputSchema`) is unchanged.
-
-This drops the `ajv` and `ajv-formats` dependencies. Interrupt binding hashes and
-resolution fingerprints now use a small bundled SHA-256 instead of
-`@noble/hashes`, so that dependency is gone too. The wire hash shape
-(`sha256:<hex>`) is unchanged.
diff --git a/.changeset/memory-middleware.md b/.changeset/memory-middleware.md
deleted file mode 100644
index 6e9cd9f31..000000000
--- a/.changeset/memory-middleware.md
+++ /dev/null
@@ -1,44 +0,0 @@
----
-'@tanstack/ai': minor
-'@tanstack/ai-event-client': minor
-'@tanstack/ai-memory': minor
----
-
-**Add server-side memory via a `recall`/`save` adapter contract in `@tanstack/ai-memory`.**
-
-Memory is now a single, provider-agnostic contract with two verbs — `recall` and
-`save` — which is the shape every memory backend (in-process, Redis, and hosted
-vendors) naturally exposes. `memoryMiddleware` recalls relevant memory into the
-system prompt (and optionally injects vendor tools) before the model runs, then
-defers `save` of the finished turn via `ctx.defer` so streaming is never blocked.
-Extraction, ranking, and rendering live inside each adapter — the middleware is thin.
-
-`@tanstack/ai-memory` (new package) — everything ships here:
-
-- Root: `memoryMiddleware`, the `MemoryAdapter` contract
-  (`recall` / `save` / optional `inspect` / `listFacts`), and the `MemoryScope` /
-  `MemoryTurn` / `RecallResult` / `SaveReceipt` types.
-- `@tanstack/ai-memory/in-memory` → `inMemory()` — zero-dependency adapter for dev,
-  tests, and single-process demos. Pass an `embedder` for semantic scoring and/or an
-  `extract` function to persist derived facts.
-- `@tanstack/ai-memory/redis` → `redis({ redis, prefix? })` — production adapter for
-  plain Redis. `ioredis` wires in directly; `redis` (node-redis v4+) via the
-  `fromNodeRedis(client)` wrapper. Both are optional peer dependencies.
-- `@tanstack/ai-memory/hindsight` → `hindsight()`, `@tanstack/ai-memory/mem0` →
-  `mem0()`, `@tanstack/ai-memory/honcho` → `honcho()` — hosted-vendor adapters. Their
-  SDKs (`@vectorize-io/hindsight-client`, `@honcho-ai/sdk`) are optional peers loaded
-  lazily; mem0 talks to its server over plain HTTP (no SDK). Vendors can expose LLM
-  tools through `recall` (e.g. hindsight's retain/recall/reflect).
-- A shared `recall`/`save` contract-test suite (`@tanstack/ai-memory/tests/contract`)
-  that any adapter — including third-party ones — can run.
-
-`@tanstack/ai`:
-
-- **Removes the (unreleased) `@tanstack/ai/memory` subpath.** The middleware,
-  contract, and helpers all moved to `@tanstack/ai-memory`.
-
-`@tanstack/ai-event-client`:
-
-- The five `memory:*` devtools events (`memory:retrieve:started` / `:completed`,
-  `memory:persist:started` / `:completed`, `memory:error`) now carry recall/save
-  payloads (adapter id, fragment/receipt counts, `phase: 'recall' | 'save'`).
diff --git a/.changeset/memory-scope-threadid.md b/.changeset/memory-scope-threadid.md
deleted file mode 100644
index ba5c66bfe..000000000
--- a/.changeset/memory-scope-threadid.md
+++ /dev/null
@@ -1,26 +0,0 @@
----
-'@tanstack/ai-memory': minor
-'@tanstack/ai-event-client': minor
-'@tanstack/ai-client': minor
-'@tanstack/ai-devtools-core': minor
----
-
-**Align `MemoryScope` to the shared `Scope` type (`threadId`).**
-
-`MemoryScope` is now an alias of `Scope` from `@tanstack/ai` so memory and
-persistence share one isolation vocabulary. The conversation key is
-`threadId` (required); optional dims are `userId`, `tenantId`, and reserved
-`namespace`. There is no public `sessionId` on memory scope — hard cut while
-`@tanstack/ai-memory` is still `0.x` / unreleased.
-
-- `@tanstack/ai-memory` — `export type MemoryScope = Scope`. Built-in adapters
-  (`inMemory`, `redis`) and middleware use `threadId`; `sameScope` also matches
-  `tenantId` when present on the query. Redis index keys are now
-  `{prefix}:index:{tenantId|_}:{userId|_}:{threadId}` (escaped). Hindsight banks
-  use `{user}__{threadId}`. Anyone who wrote Redis rows under the pre-rename
-  layout needs to reindex or wipe — keys are not dual-read.
-- `@tanstack/ai-event-client` — `MemoryScopeLite` is
-  `{ threadId?, userId?, tenantId? }` (devtools telemetry; not an isolation
-  authority).
-- `@tanstack/ai-client` / `@tanstack/ai-devtools-core` — memory event payloads
-  and the Memory panel registry follow the same `threadId` field names.
diff --git a/.changeset/resumable-streams.md b/.changeset/resumable-streams.md
deleted file mode 100644
index 35b6ebfd7..000000000
--- a/.changeset/resumable-streams.md
+++ /dev/null
@@ -1,53 +0,0 @@
----
-'@tanstack/ai': minor
-'@tanstack/ai-client': minor
-'@tanstack/ai-durable-stream': minor
----
-
-Resumable streams: reconnect to an in-flight SSE **or NDJSON** response without
-re-running the provider.
-
-`toServerSentEventsResponse` and `toHttpResponse` both accept a
-`durability: { adapter, batch }` option. The adapter (`StreamDurability`)
-records every chunk to an ordered log before delivery and tags each event with
-an opaque, adapter-owned offset — an SSE `id:` line, or the `id` of an NDJSON
-`{ id, chunk }` envelope (NDJSON has no native event-id). A reconnect
-(`Last-Event-ID`) or an explicit `?offset` read replays strictly after that
-offset from the log — the lazy provider stream is never iterated on resume.
-Producers terminalize the log on cancellation and failure (`RUN_ERROR` append
-
-- `close()`) and on completion when the source stream emits its own terminal
-  event (`chat()` always does), so readers are never parked on a dead run.
-
-Two adapters ship: `memoryStream(request)` in `@tanstack/ai` (process-local,
-for development and tests) and the new `@tanstack/ai-durable-stream` package,
-a Durable Streams protocol adapter for production backends.
-
-For the `GET` handler that a reload or a second tab reconnects to,
-`resumeServerSentEventsResponse({ adapter })` and `resumeHttpResponse({ adapter })`
-replay a run straight from the durability log. They need no producer stream and
-return a 400 when the request carries no resume offset.
-
-On the client, all four HTTP adapters are now resumable — `fetchServerSentEvents`,
-`fetchHttpStream`, `xhrServerSentEvents`, and `xhrHttpStream`. Each tracks the
-per-event offset, auto-reconnects with `Last-Event-ID`, de-duplicates the
-replayed prefix, and exposes `joinRun(runId)` to attach to an in-flight or
-finished run from the start (read-only GET with `offset=-1`). Untagged streams
-behave exactly as before. A durable run that ends with no terminal event and no
-forward progress now throws `DurableStreamIncompleteError` instead of hanging.
-
-Reconnection and durability are bounded so failures surface rather than hang or
-loop:
-
-- `memoryStream` evicts completed logs after a grace window (unbounded growth
-  is gone); resuming an expired/unknown run throws, and a from-start join to a
-  run that never produces fails after `MemoryStreamOptions.firstChunkDeadlineMs`.
-- all four HTTP adapters accept `reconnect: { maxAttempts, delayMs }` — a
-  throttle plus a ceiling on CONSECUTIVE no-progress reconnects (default 5;
-  forward progress resets it) that fails with the new `StreamReconnectLimitError`
-  instead of reconnecting endlessly, without penalizing a healthy long-lived run.
-- `durableStream` accepts `reconnect: { maxReadFailures, delayMs }` to bound its
-  read-retry loop, and `server` is now optional when `fetch` is provided (e.g. a
-  Cloudflare service binding).
-- `toServerSentEventsResponse` accepts `debug` to record durability terminal /
-  close failures server-side, where a replaying joiner cannot observe them.
diff --git a/.changeset/shared-scope.md b/.changeset/shared-scope.md
deleted file mode 100644
index 7dee65362..000000000
--- a/.changeset/shared-scope.md
+++ /dev/null
@@ -1,29 +0,0 @@
----
-'@tanstack/ai': minor
----
-
-**Add a shared `Scope` identity type to `@tanstack/ai`.**
-
-`Scope` is the single identity/isolation vocabulary for the subsystems that
-persist or recall per-conversation data — `@tanstack/ai-persistence` and
-`@tanstack/ai-memory`. Rather than each subsystem inventing its own notion of
-"whose data is this?", both import one type:
-
-```ts
-interface Scope {
-  threadId: string // required — the single conversation key (same as ctx.threadId)
-  userId?: string // durable end-user identity; required in practice for multi-user apps
-  tenantId?: string // multi-tenant boundary
-  namespace?: string // reserved logical partition; no subsystem keys on it yet
-}
-```
-
-`threadId` is the one conversation key across the codebase (matching
-`ChatMiddlewareContext.threadId`, with `conversationId` already deprecated in
-favor of it) — subsystems must not introduce a second name (`sessionId`, …) for
-the same concept. Every field is an isolation boundary and must be derived
-server-side from trusted session state, never from client input.
-
-Introduced ahead of the persistence and memory packages so both share one settled
-identity contract. `@tanstack/ai-memory` now aliases `MemoryScope` to `Scope`
-(see the memory-scope-threadid changeset).
diff --git a/.changeset/sync-models.md b/.changeset/sync-models.md
deleted file mode 100644
index 63eef764c..000000000
--- a/.changeset/sync-models.md
+++ /dev/null
@@ -1,5 +0,0 @@
----
-'@tanstack/ai-openrouter': patch
----
-
-Update model metadata from OpenRouter API
diff --git a/examples/ag-ui/CHANGELOG.md b/examples/ag-ui/CHANGELOG.md
index 41d1e2e02..fbd5a7bd2 100644
--- a/examples/ag-ui/CHANGELOG.md
+++ b/examples/ag-ui/CHANGELOG.md
@@ -1,5 +1,14 @@
 # ag-ui
 
+## 0.0.3
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288)]:
+  - @tanstack/ai-client@0.23.0
+  - @tanstack/ai-react@0.19.0
+  - @tanstack/ai-react-ui@0.8.16
+
 ## 0.0.2
 
 ### Patch Changes
diff --git a/examples/ag-ui/package.json b/examples/ag-ui/package.json
index 3475ab6f6..2eb3f2b18 100644
--- a/examples/ag-ui/package.json
+++ b/examples/ag-ui/package.json
@@ -2,7 +2,7 @@
   "name": "ag-ui",
   "private": true,
   "type": "module",
-  "version": "0.0.2",
+  "version": "0.0.3",
   "scripts": {
     "detect-servers": "node scripts/detect-servers.mjs",
     "predev": "node scripts/detect-servers.mjs",
diff --git a/packages/ai-acp/CHANGELOG.md b/packages/ai-acp/CHANGELOG.md
index 663b6797f..c308530dd 100644
--- a/packages/ai-acp/CHANGELOG.md
+++ b/packages/ai-acp/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-acp
 
+## 0.2.4
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-sandbox@0.2.5
+
 ## 0.2.3
 
 ### Patch Changes
diff --git a/packages/ai-acp/package.json b/packages/ai-acp/package.json
index 7f766d480..dcb042296 100644
--- a/packages/ai-acp/package.json
+++ b/packages/ai-acp/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-acp",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "description": "Shared Agent Client Protocol (ACP) transport, session, and AG-UI translation for TanStack AI harness adapters.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-angular/CHANGELOG.md b/packages/ai-angular/CHANGELOG.md
index 32645461d..aa2fd4324 100644
--- a/packages/ai-angular/CHANGELOG.md
+++ b/packages/ai-angular/CHANGELOG.md
@@ -1,5 +1,26 @@
 # @tanstack/ai-angular
 
+## 0.4.0
+
+### Minor Changes
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
+  client-tool execution, with typed bound resolvers, atomic batches, and
+  structured errors. Interrupts run ephemerally by resuming from the full client
+  message history in a fresh child run — no persistence required.
+
+  This changes native approval and client-tool streams from legacy custom events
+  to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
+  `pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
+  legacy event readers remain as limited compatibility surfaces for migration;
+  `addToolResult` remains supported.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-client@0.23.0
+
 ## 0.3.1
 
 ### Patch Changes
diff --git a/packages/ai-angular/package.json b/packages/ai-angular/package.json
index 0cc622277..fb1d48aa9 100644
--- a/packages/ai-angular/package.json
+++ b/packages/ai-angular/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-angular",
-  "version": "0.3.1",
+  "version": "0.4.0",
   "description": "Angular signals integration for TanStack AI streaming chat, structured outputs, and media generation.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-anthropic/CHANGELOG.md b/packages/ai-anthropic/CHANGELOG.md
index f9f654410..ec8d5d0d1 100644
--- a/packages/ai-anthropic/CHANGELOG.md
+++ b/packages/ai-anthropic/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-anthropic
 
+## 0.16.4
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.16.3
 
 ### Patch Changes
diff --git a/packages/ai-anthropic/package.json b/packages/ai-anthropic/package.json
index 4f6702cbb..ba67655c8 100644
--- a/packages/ai-anthropic/package.json
+++ b/packages/ai-anthropic/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-anthropic",
-  "version": "0.16.3",
+  "version": "0.16.4",
   "description": "Anthropic Claude adapter for TanStack AI chat, tool calling, thinking, and structured outputs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-bedrock/CHANGELOG.md b/packages/ai-bedrock/CHANGELOG.md
index d737219ad..9aa8a5557 100644
--- a/packages/ai-bedrock/CHANGELOG.md
+++ b/packages/ai-bedrock/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-bedrock
 
+## 0.1.5
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/openai-base@0.9.10
+
 ## 0.1.4
 
 ### Patch Changes
diff --git a/packages/ai-bedrock/package.json b/packages/ai-bedrock/package.json
index 1981b2b4c..5ac0ae8fa 100644
--- a/packages/ai-bedrock/package.json
+++ b/packages/ai-bedrock/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-bedrock",
-  "version": "0.1.4",
+  "version": "0.1.5",
   "type": "module",
   "description": "Amazon Bedrock adapter for TanStack AI — OpenAI-compatible chat, responses, tools, and reasoning.",
   "author": "",
diff --git a/packages/ai-claude-code/CHANGELOG.md b/packages/ai-claude-code/CHANGELOG.md
index 048c0b7d7..bdedcc5cc 100644
--- a/packages/ai-claude-code/CHANGELOG.md
+++ b/packages/ai-claude-code/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-claude-code
 
+## 0.2.4
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-sandbox@0.2.5
+
 ## 0.2.3
 
 ### Patch Changes
diff --git a/packages/ai-claude-code/package.json b/packages/ai-claude-code/package.json
index 9d960189c..fa976a33f 100644
--- a/packages/ai-claude-code/package.json
+++ b/packages/ai-claude-code/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-claude-code",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "description": "Claude Code harness adapter for TanStack AI — run Claude Code as a chat backend with local tool execution and stateful sessions.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-client/CHANGELOG.md b/packages/ai-client/CHANGELOG.md
index e0b56d32f..100cad67d 100644
--- a/packages/ai-client/CHANGELOG.md
+++ b/packages/ai-client/CHANGELOG.md
@@ -1,5 +1,157 @@
 # @tanstack/ai-client
 
+## 0.23.0
+
+### Minor Changes
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
+  client-tool execution, with typed bound resolvers, atomic batches, and
+  structured errors. Interrupts run ephemerally by resuming from the full client
+  message history in a fresh child run — no persistence required.
+
+  This changes native approval and client-tool streams from legacy custom events
+  to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
+  `pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
+  legacy event readers remain as limited compatibility surfaces for migration;
+  `addToolResult` remains supported.
+
+- [#541](https://github.com/TanStack/ai/pull/541) [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4) - **Surface server-side memory state in the TanStack AI DevTools.**
+
+  The DevTools panel now has a **Memory** tab for any chat wired with
+  `memoryMiddleware`. It shows, per scope (session), an operations timeline (each
+  turn's recall — query, fragment count, injected system-prompt size, whether
+  memory tools were exposed, duration) and the current stored records/facts when
+  the adapter implements the optional `inspect`/`listFacts` methods.
+
+  Because memory runs on the server (whose event bus never reaches the browser),
+  the middleware transports its state to the panel over the chat stream as a
+  `memory:state` `CUSTOM` event, which `@tanstack/ai-client`'s devtools bridge
+  re-emits as browser `memory:*` events — the same pattern generation results use.
+  The snapshot reflects memory as of the start of each turn; opening the panel
+  mid-conversation replays the latest state so the tab isn't empty.
+  - `@tanstack/ai-memory` — `memoryMiddleware` injects a `memory:state` `CUSTOM`
+    chunk carrying recall metrics + an `inspect`/`listFacts` snapshot; exports
+    `MEMORY_STATE_EVENT` and `MemoryStateEventValue`.
+  - `@tanstack/ai-event-client` — adds the `memory:snapshot` devtools event.
+  - `@tanstack/ai-client` — the chat devtools bridge re-emits `memory:*` from the
+    transported chunk and replays the last snapshot on `devtools:request-state`.
+  - `@tanstack/ai-devtools-core` — new Memory tab + per-scope memory store slice.
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Make interrupt ownership explicit rather than assumed.
+
+  An AG-UI `Interrupt` is a shared envelope — a workflow engine's durable
+  approval or another agent framework's pause can arrive on the same stream. What
+  makes a pause resumable through `chat()` is the binding this package attaches
+  under `tanstack:interruptBinding`.
+  - Interrupts that carry no binding this client understands now surface as
+    `kind: 'unbound'` with `canResolve: false`, instead of being given a
+    synthesized binding and rendered as resolvable generic interrupts. Resolving
+    those produced an answer submitted against a run with nothing pending, which
+    failed as `unknown-interrupt` only after the user had filled in the form.
+    Unbound items never block submission of the interrupts that are yours.
+  - The binding carries a wire version (`INTERRUPT_BINDING_VERSION`). Readers
+    reject a version they don't recognise rather than duck-typing its fields. A
+    binding written before the field existed is still read.
+  - `INTERRUPT_BINDING_METADATA_KEY`, `withInterruptBinding()` and
+    `readInterruptBinding()` are exported, so anything producing an interrupt this
+    package must later resume attaches the binding through a supported API
+    instead of copying the metadata key.
+  - Interrupt classification is driven by the binding alone. `Interrupt.reason` is
+    free-form AG-UI text another producer can also use, so it is now a display
+    hint only and never decides ownership.
+  - The interrupt protocol surface is enumerated instead of `export *`. The
+    unimplemented durable-recovery contract (`InterruptRecoveryStateV1`,
+    `InterruptRecoveryQuery`, the never-called `loadInterruptState` adapter hook,
+    and the `persistence-required` / `atomic-commit-unsupported` /
+    `recovery-unavailable` error codes) is removed rather than published.
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Interrupts: the application owns wire-schema validation, and the hashing
+  dependency is gone.
+
+  The library no longer transforms a generic interrupt's wire JSON Schema into a
+  validator or validates the resolved value against it, on either the client or
+  the server. Whatever you pass to `resolveInterrupt` (client) or send in the
+  `resume` batch (server) flows through as-is. Validate it yourself if you need to
+  trust it, e.g. with `z.fromJSONSchema(interrupt.responseSchema).safeParse(value)`
+  on the client and your own check on the server. Validation of a tool's
+  code-authored Standard Schema (`approvalSchema` / `inputSchema`) is unchanged.
+
+  This drops the `ajv` and `ajv-formats` dependencies. Interrupt binding hashes and
+  resolution fingerprints now use a small bundled SHA-256 instead of
+  `@noble/hashes`, so that dependency is gone too. The wire hash shape
+  (`sha256:<hex>`) is unchanged.
+
+- [#991](https://github.com/TanStack/ai/pull/991) [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371) - **Align `MemoryScope` to the shared `Scope` type (`threadId`).**
+
+  `MemoryScope` is now an alias of `Scope` from `@tanstack/ai` so memory and
+  persistence share one isolation vocabulary. The conversation key is
+  `threadId` (required); optional dims are `userId`, `tenantId`, and reserved
+  `namespace`. There is no public `sessionId` on memory scope — hard cut while
+  `@tanstack/ai-memory` is still `0.x` / unreleased.
+  - `@tanstack/ai-memory` — `export type MemoryScope = Scope`. Built-in adapters
+    (`inMemory`, `redis`) and middleware use `threadId`; `sameScope` also matches
+    `tenantId` when present on the query. Redis index keys are now
+    `{prefix}:index:{tenantId|_}:{userId|_}:{threadId}` (escaped). Hindsight banks
+    use `{user}__{threadId}`. Anyone who wrote Redis rows under the pre-rename
+    layout needs to reindex or wipe — keys are not dual-read.
+  - `@tanstack/ai-event-client` — `MemoryScopeLite` is
+    `{ threadId?, userId?, tenantId? }` (devtools telemetry; not an isolation
+    authority).
+  - `@tanstack/ai-client` / `@tanstack/ai-devtools-core` — memory event payloads
+    and the Memory panel registry follow the same `threadId` field names.
+
+- [#955](https://github.com/TanStack/ai/pull/955) [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288) - Resumable streams: reconnect to an in-flight SSE **or NDJSON** response without
+  re-running the provider.
+
+  `toServerSentEventsResponse` and `toHttpResponse` both accept a
+  `durability: { adapter, batch }` option. The adapter (`StreamDurability`)
+  records every chunk to an ordered log before delivery and tags each event with
+  an opaque, adapter-owned offset — an SSE `id:` line, or the `id` of an NDJSON
+  `{ id, chunk }` envelope (NDJSON has no native event-id). A reconnect
+  (`Last-Event-ID`) or an explicit `?offset` read replays strictly after that
+  offset from the log — the lazy provider stream is never iterated on resume.
+  Producers terminalize the log on cancellation and failure (`RUN_ERROR` append
+  - `close()`) and on completion when the source stream emits its own terminal
+    event (`chat()` always does), so readers are never parked on a dead run.
+
+  Two adapters ship: `memoryStream(request)` in `@tanstack/ai` (process-local,
+  for development and tests) and the new `@tanstack/ai-durable-stream` package,
+  a Durable Streams protocol adapter for production backends.
+
+  For the `GET` handler that a reload or a second tab reconnects to,
+  `resumeServerSentEventsResponse({ adapter })` and `resumeHttpResponse({ adapter })`
+  replay a run straight from the durability log. They need no producer stream and
+  return a 400 when the request carries no resume offset.
+
+  On the client, all four HTTP adapters are now resumable — `fetchServerSentEvents`,
+  `fetchHttpStream`, `xhrServerSentEvents`, and `xhrHttpStream`. Each tracks the
+  per-event offset, auto-reconnects with `Last-Event-ID`, de-duplicates the
+  replayed prefix, and exposes `joinRun(runId)` to attach to an in-flight or
+  finished run from the start (read-only GET with `offset=-1`). Untagged streams
+  behave exactly as before. A durable run that ends with no terminal event and no
+  forward progress now throws `DurableStreamIncompleteError` instead of hanging.
+
+  Reconnection and durability are bounded so failures surface rather than hang or
+  loop:
+  - `memoryStream` evicts completed logs after a grace window (unbounded growth
+    is gone); resuming an expired/unknown run throws, and a from-start join to a
+    run that never produces fails after `MemoryStreamOptions.firstChunkDeadlineMs`.
+  - all four HTTP adapters accept `reconnect: { maxAttempts, delayMs }` — a
+    throttle plus a ceiling on CONSECUTIVE no-progress reconnects (default 5;
+    forward progress resets it) that fails with the new `StreamReconnectLimitError`
+    instead of reconnecting endlessly, without penalizing a healthy long-lived run.
+  - `durableStream` accepts `reconnect: { maxReadFailures, delayMs }` to bound its
+    read-retry loop, and `server` is now optional when `fetch` is provided (e.g. a
+    Cloudflare service binding).
+  - `toServerSentEventsResponse` accepts `debug` to record durability terminal /
+    close failures server-side, where a replaying joiner cannot observe them.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-event-client@0.7.0
+
 ## 0.22.1
 
 ### Patch Changes
diff --git a/packages/ai-client/package.json b/packages/ai-client/package.json
index 2423f6948..1d11e6141 100644
--- a/packages/ai-client/package.json
+++ b/packages/ai-client/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-client",
-  "version": "0.22.1",
+  "version": "0.23.0",
   "description": "Framework-agnostic headless client for TanStack AI chat, realtime sessions, streaming transports, and media generations.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-code-mode-skills/CHANGELOG.md b/packages/ai-code-mode-skills/CHANGELOG.md
index 94c22b6a9..055a758e4 100644
--- a/packages/ai-code-mode-skills/CHANGELOG.md
+++ b/packages/ai-code-mode-skills/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-code-mode-skills
 
+## 0.3.12
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-code-mode@0.3.9
+
 ## 0.3.11
 
 ### Patch Changes
diff --git a/packages/ai-code-mode-skills/package.json b/packages/ai-code-mode-skills/package.json
index 48ffd8d90..64782b2f7 100644
--- a/packages/ai-code-mode-skills/package.json
+++ b/packages/ai-code-mode-skills/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-code-mode-skills",
-  "version": "0.3.11",
+  "version": "0.3.12",
   "description": "Persistent runtime skill library for TanStack AI Code Mode agents and sandboxed tool orchestration.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-code-mode/CHANGELOG.md b/packages/ai-code-mode/CHANGELOG.md
index 3ca87fd9d..03c7cf574 100644
--- a/packages/ai-code-mode/CHANGELOG.md
+++ b/packages/ai-code-mode/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-code-mode
 
+## 0.3.9
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.3.8
 
 ### Patch Changes
diff --git a/packages/ai-code-mode/package.json b/packages/ai-code-mode/package.json
index 85a837d0b..90fb51501 100644
--- a/packages/ai-code-mode/package.json
+++ b/packages/ai-code-mode/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-code-mode",
-  "version": "0.3.8",
+  "version": "0.3.9",
   "description": "Secure TypeScript Code Mode for TanStack AI agents to execute sandboxed tool orchestration programs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-codex/CHANGELOG.md b/packages/ai-codex/CHANGELOG.md
index 6710f31d7..be9dd021d 100644
--- a/packages/ai-codex/CHANGELOG.md
+++ b/packages/ai-codex/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-codex
 
+## 0.2.4
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-sandbox@0.2.5
+
 ## 0.2.3
 
 ### Patch Changes
diff --git a/packages/ai-codex/package.json b/packages/ai-codex/package.json
index 4bdfad50c..2ba51c787 100644
--- a/packages/ai-codex/package.json
+++ b/packages/ai-codex/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-codex",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "description": "Codex harness adapter for TanStack AI — run OpenAI Codex as a chat backend with local tool execution and stateful sessions.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-devtools/CHANGELOG.md b/packages/ai-devtools/CHANGELOG.md
index 9a00d78c5..b13210918 100644
--- a/packages/ai-devtools/CHANGELOG.md
+++ b/packages/ai-devtools/CHANGELOG.md
@@ -1,5 +1,56 @@
 # @tanstack/ai-devtools-core
 
+## 0.5.0
+
+### Minor Changes
+
+- [#541](https://github.com/TanStack/ai/pull/541) [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4) - **Surface server-side memory state in the TanStack AI DevTools.**
+
+  The DevTools panel now has a **Memory** tab for any chat wired with
+  `memoryMiddleware`. It shows, per scope (session), an operations timeline (each
+  turn's recall — query, fragment count, injected system-prompt size, whether
+  memory tools were exposed, duration) and the current stored records/facts when
+  the adapter implements the optional `inspect`/`listFacts` methods.
+
+  Because memory runs on the server (whose event bus never reaches the browser),
+  the middleware transports its state to the panel over the chat stream as a
+  `memory:state` `CUSTOM` event, which `@tanstack/ai-client`'s devtools bridge
+  re-emits as browser `memory:*` events — the same pattern generation results use.
+  The snapshot reflects memory as of the start of each turn; opening the panel
+  mid-conversation replays the latest state so the tab isn't empty.
+  - `@tanstack/ai-memory` — `memoryMiddleware` injects a `memory:state` `CUSTOM`
+    chunk carrying recall metrics + an `inspect`/`listFacts` snapshot; exports
+    `MEMORY_STATE_EVENT` and `MemoryStateEventValue`.
+  - `@tanstack/ai-event-client` — adds the `memory:snapshot` devtools event.
+  - `@tanstack/ai-client` — the chat devtools bridge re-emits `memory:*` from the
+    transported chunk and replays the last snapshot on `devtools:request-state`.
+  - `@tanstack/ai-devtools-core` — new Memory tab + per-scope memory store slice.
+
+- [#991](https://github.com/TanStack/ai/pull/991) [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371) - **Align `MemoryScope` to the shared `Scope` type (`threadId`).**
+
+  `MemoryScope` is now an alias of `Scope` from `@tanstack/ai` so memory and
+  persistence share one isolation vocabulary. The conversation key is
+  `threadId` (required); optional dims are `userId`, `tenantId`, and reserved
+  `namespace`. There is no public `sessionId` on memory scope — hard cut while
+  `@tanstack/ai-memory` is still `0.x` / unreleased.
+  - `@tanstack/ai-memory` — `export type MemoryScope = Scope`. Built-in adapters
+    (`inMemory`, `redis`) and middleware use `threadId`; `sameScope` also matches
+    `tenantId` when present on the query. Redis index keys are now
+    `{prefix}:index:{tenantId|_}:{userId|_}:{threadId}` (escaped). Hindsight banks
+    use `{user}__{threadId}`. Anyone who wrote Redis rows under the pre-rename
+    layout needs to reindex or wipe — keys are not dual-read.
+  - `@tanstack/ai-event-client` — `MemoryScopeLite` is
+    `{ threadId?, userId?, tenantId? }` (devtools telemetry; not an isolation
+    authority).
+  - `@tanstack/ai-client` / `@tanstack/ai-devtools-core` — memory event payloads
+    and the Memory panel registry follow the same `threadId` field names.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-event-client@0.7.0
+
 ## 0.4.24
 
 ### Patch Changes
diff --git a/packages/ai-devtools/package.json b/packages/ai-devtools/package.json
index e0a766d97..84b3a0452 100644
--- a/packages/ai-devtools/package.json
+++ b/packages/ai-devtools/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-devtools-core",
-  "version": "0.4.24",
+  "version": "0.5.0",
   "description": "Core TanStack AI Devtools plugin for inspecting chat messages, tool calls, streams, and errors.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-durable-stream/CHANGELOG.md b/packages/ai-durable-stream/CHANGELOG.md
new file mode 100644
index 000000000..9f16f841e
--- /dev/null
+++ b/packages/ai-durable-stream/CHANGELOG.md
@@ -0,0 +1,56 @@
+# @tanstack/ai-durable-stream
+
+## 0.1.0
+
+### Minor Changes
+
+- [#955](https://github.com/TanStack/ai/pull/955) [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288) - Resumable streams: reconnect to an in-flight SSE **or NDJSON** response without
+  re-running the provider.
+
+  `toServerSentEventsResponse` and `toHttpResponse` both accept a
+  `durability: { adapter, batch }` option. The adapter (`StreamDurability`)
+  records every chunk to an ordered log before delivery and tags each event with
+  an opaque, adapter-owned offset — an SSE `id:` line, or the `id` of an NDJSON
+  `{ id, chunk }` envelope (NDJSON has no native event-id). A reconnect
+  (`Last-Event-ID`) or an explicit `?offset` read replays strictly after that
+  offset from the log — the lazy provider stream is never iterated on resume.
+  Producers terminalize the log on cancellation and failure (`RUN_ERROR` append
+  - `close()`) and on completion when the source stream emits its own terminal
+    event (`chat()` always does), so readers are never parked on a dead run.
+
+  Two adapters ship: `memoryStream(request)` in `@tanstack/ai` (process-local,
+  for development and tests) and the new `@tanstack/ai-durable-stream` package,
+  a Durable Streams protocol adapter for production backends.
+
+  For the `GET` handler that a reload or a second tab reconnects to,
+  `resumeServerSentEventsResponse({ adapter })` and `resumeHttpResponse({ adapter })`
+  replay a run straight from the durability log. They need no producer stream and
+  return a 400 when the request carries no resume offset.
+
+  On the client, all four HTTP adapters are now resumable — `fetchServerSentEvents`,
+  `fetchHttpStream`, `xhrServerSentEvents`, and `xhrHttpStream`. Each tracks the
+  per-event offset, auto-reconnects with `Last-Event-ID`, de-duplicates the
+  replayed prefix, and exposes `joinRun(runId)` to attach to an in-flight or
+  finished run from the start (read-only GET with `offset=-1`). Untagged streams
+  behave exactly as before. A durable run that ends with no terminal event and no
+  forward progress now throws `DurableStreamIncompleteError` instead of hanging.
+
+  Reconnection and durability are bounded so failures surface rather than hang or
+  loop:
+  - `memoryStream` evicts completed logs after a grace window (unbounded growth
+    is gone); resuming an expired/unknown run throws, and a from-start join to a
+    run that never produces fails after `MemoryStreamOptions.firstChunkDeadlineMs`.
+  - all four HTTP adapters accept `reconnect: { maxAttempts, delayMs }` — a
+    throttle plus a ceiling on CONSECUTIVE no-progress reconnects (default 5;
+    forward progress resets it) that fails with the new `StreamReconnectLimitError`
+    instead of reconnecting endlessly, without penalizing a healthy long-lived run.
+  - `durableStream` accepts `reconnect: { maxReadFailures, delayMs }` to bound its
+    read-retry loop, and `server` is now optional when `fetch` is provided (e.g. a
+    Cloudflare service binding).
+  - `toServerSentEventsResponse` accepts `debug` to record durability terminal /
+    close failures server-side, where a replaying joiner cannot observe them.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
diff --git a/packages/ai-durable-stream/package.json b/packages/ai-durable-stream/package.json
index 2ce39ffb8..dd3b365fb 100644
--- a/packages/ai-durable-stream/package.json
+++ b/packages/ai-durable-stream/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-durable-stream",
-  "version": "0.0.0",
+  "version": "0.1.0",
   "description": "Delivery durability for TanStack AI over the durable-streams HTTP protocol — a resumable StreamDurability transport sink (append/read/resume) that stores zero delivery events itself.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-elevenlabs/CHANGELOG.md b/packages/ai-elevenlabs/CHANGELOG.md
index e441d6d16..7361afebd 100644
--- a/packages/ai-elevenlabs/CHANGELOG.md
+++ b/packages/ai-elevenlabs/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-elevenlabs
 
+## 0.2.35
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.2.34
 
 ### Patch Changes
diff --git a/packages/ai-elevenlabs/package.json b/packages/ai-elevenlabs/package.json
index ee1c39d20..d4d7e6648 100644
--- a/packages/ai-elevenlabs/package.json
+++ b/packages/ai-elevenlabs/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-elevenlabs",
-  "version": "0.2.34",
+  "version": "0.2.35",
   "description": "ElevenLabs adapter for TanStack AI realtime voice, text-to-speech, transcription, music, and sound effects.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-event-client/CHANGELOG.md b/packages/ai-event-client/CHANGELOG.md
index 636549768..b70153785 100644
--- a/packages/ai-event-client/CHANGELOG.md
+++ b/packages/ai-event-client/CHANGELOG.md
@@ -1,5 +1,86 @@
 # @tanstack/ai-event-client
 
+## 0.7.0
+
+### Minor Changes
+
+- [#541](https://github.com/TanStack/ai/pull/541) [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4) - **Surface server-side memory state in the TanStack AI DevTools.**
+
+  The DevTools panel now has a **Memory** tab for any chat wired with
+  `memoryMiddleware`. It shows, per scope (session), an operations timeline (each
+  turn's recall — query, fragment count, injected system-prompt size, whether
+  memory tools were exposed, duration) and the current stored records/facts when
+  the adapter implements the optional `inspect`/`listFacts` methods.
+
+  Because memory runs on the server (whose event bus never reaches the browser),
+  the middleware transports its state to the panel over the chat stream as a
+  `memory:state` `CUSTOM` event, which `@tanstack/ai-client`'s devtools bridge
+  re-emits as browser `memory:*` events — the same pattern generation results use.
+  The snapshot reflects memory as of the start of each turn; opening the panel
+  mid-conversation replays the latest state so the tab isn't empty.
+  - `@tanstack/ai-memory` — `memoryMiddleware` injects a `memory:state` `CUSTOM`
+    chunk carrying recall metrics + an `inspect`/`listFacts` snapshot; exports
+    `MEMORY_STATE_EVENT` and `MemoryStateEventValue`.
+  - `@tanstack/ai-event-client` — adds the `memory:snapshot` devtools event.
+  - `@tanstack/ai-client` — the chat devtools bridge re-emits `memory:*` from the
+    transported chunk and replays the last snapshot on `devtools:request-state`.
+  - `@tanstack/ai-devtools-core` — new Memory tab + per-scope memory store slice.
+
+- [#541](https://github.com/TanStack/ai/pull/541) [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4) - **Add server-side memory via a `recall`/`save` adapter contract in `@tanstack/ai-memory`.**
+
+  Memory is now a single, provider-agnostic contract with two verbs — `recall` and
+  `save` — which is the shape every memory backend (in-process, Redis, and hosted
+  vendors) naturally exposes. `memoryMiddleware` recalls relevant memory into the
+  system prompt (and optionally injects vendor tools) before the model runs, then
+  defers `save` of the finished turn via `ctx.defer` so streaming is never blocked.
+  Extraction, ranking, and rendering live inside each adapter — the middleware is thin.
+
+  `@tanstack/ai-memory` (new package) — everything ships here:
+  - Root: `memoryMiddleware`, the `MemoryAdapter` contract
+    (`recall` / `save` / optional `inspect` / `listFacts`), and the `MemoryScope` /
+    `MemoryTurn` / `RecallResult` / `SaveReceipt` types.
+  - `@tanstack/ai-memory/in-memory` → `inMemory()` — zero-dependency adapter for dev,
+    tests, and single-process demos. Pass an `embedder` for semantic scoring and/or an
+    `extract` function to persist derived facts.
+  - `@tanstack/ai-memory/redis` → `redis({ redis, prefix? })` — production adapter for
+    plain Redis. `ioredis` wires in directly; `redis` (node-redis v4+) via the
+    `fromNodeRedis(client)` wrapper. Both are optional peer dependencies.
+  - `@tanstack/ai-memory/hindsight` → `hindsight()`, `@tanstack/ai-memory/mem0` →
+    `mem0()`, `@tanstack/ai-memory/honcho` → `honcho()` — hosted-vendor adapters. Their
+    SDKs (`@vectorize-io/hindsight-client`, `@honcho-ai/sdk`) are optional peers loaded
+    lazily; mem0 talks to its server over plain HTTP (no SDK). Vendors can expose LLM
+    tools through `recall` (e.g. hindsight's retain/recall/reflect).
+  - A shared `recall`/`save` contract-test suite (`@tanstack/ai-memory/tests/contract`)
+    that any adapter — including third-party ones — can run.
+
+  `@tanstack/ai`:
+  - **Removes the (unreleased) `@tanstack/ai/memory` subpath.** The middleware,
+    contract, and helpers all moved to `@tanstack/ai-memory`.
+
+  `@tanstack/ai-event-client`:
+  - The five `memory:*` devtools events (`memory:retrieve:started` / `:completed`,
+    `memory:persist:started` / `:completed`, `memory:error`) now carry recall/save
+    payloads (adapter id, fragment/receipt counts, `phase: 'recall' | 'save'`).
+
+- [#991](https://github.com/TanStack/ai/pull/991) [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371) - **Align `MemoryScope` to the shared `Scope` type (`threadId`).**
+
+  `MemoryScope` is now an alias of `Scope` from `@tanstack/ai` so memory and
+  persistence share one isolation vocabulary. The conversation key is
+  `threadId` (required); optional dims are `userId`, `tenantId`, and reserved
+  `namespace`. There is no public `sessionId` on memory scope — hard cut while
+  `@tanstack/ai-memory` is still `0.x` / unreleased.
+  - `@tanstack/ai-memory` — `export type MemoryScope = Scope`. Built-in adapters
+    (`inMemory`, `redis`) and middleware use `threadId`; `sameScope` also matches
+    `tenantId` when present on the query. Redis index keys are now
+    `{prefix}:index:{tenantId|_}:{userId|_}:{threadId}` (escaped). Hindsight banks
+    use `{user}__{threadId}`. Anyone who wrote Redis rows under the pre-rename
+    layout needs to reindex or wipe — keys are not dual-read.
+  - `@tanstack/ai-event-client` — `MemoryScopeLite` is
+    `{ threadId?, userId?, tenantId? }` (devtools telemetry; not an isolation
+    authority).
+  - `@tanstack/ai-client` / `@tanstack/ai-devtools-core` — memory event payloads
+    and the Memory panel registry follow the same `threadId` field names.
+
 ## 0.6.8
 
 ### Patch Changes
diff --git a/packages/ai-event-client/package.json b/packages/ai-event-client/package.json
index 8e7ed2a49..0a4abacc2 100644
--- a/packages/ai-event-client/package.json
+++ b/packages/ai-event-client/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-event-client",
-  "version": "0.6.8",
+  "version": "0.7.0",
   "description": "Typed event client for TanStack AI devtools, observability, and streamed runtime events.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-fal/CHANGELOG.md b/packages/ai-fal/CHANGELOG.md
index e7239bcc4..e54b089b8 100644
--- a/packages/ai-fal/CHANGELOG.md
+++ b/packages/ai-fal/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-fal
 
+## 0.9.13
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.9.12
 
 ### Patch Changes
diff --git a/packages/ai-fal/package.json b/packages/ai-fal/package.json
index 8290805ba..bb7e6b134 100644
--- a/packages/ai-fal/package.json
+++ b/packages/ai-fal/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-fal",
-  "version": "0.9.12",
+  "version": "0.9.13",
   "description": "fal.ai adapter for TanStack AI image, video, audio, speech, and transcription generation.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-gemini/CHANGELOG.md b/packages/ai-gemini/CHANGELOG.md
index 7c2f00957..3284b4af5 100644
--- a/packages/ai-gemini/CHANGELOG.md
+++ b/packages/ai-gemini/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-gemini
 
+## 0.20.2
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.20.1
 
 ### Patch Changes
@@ -22,7 +29,9 @@
   ```ts
   createOpenaiImage('gpt-image-1', apiKey, { allowUrlFetch: true })
   createOpenaiVideo('sora-2', apiKey, { allowUrlFetch: true })
-  createGeminiVideo('veo-3.1-generate-preview', apiKey, { allowUrlFetch: true })
+  createGeminiVideo('veo-3.1-generate-preview', apiKey, {
+    allowUrlFetch: true,
+  })
   ```
 
   Migration: if you passed HTTP(S) URL image inputs to these adapters, either fetch the bytes yourself and pass a `data:` URI, pass a `gs://` reference (Veo), or set `allowUrlFetch: true`.
diff --git a/packages/ai-gemini/package.json b/packages/ai-gemini/package.json
index 412fa0a58..e7cb5e9a5 100644
--- a/packages/ai-gemini/package.json
+++ b/packages/ai-gemini/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-gemini",
-  "version": "0.20.1",
+  "version": "0.20.2",
   "description": "Google Gemini adapter for TanStack AI chat, images, speech, audio generation, and structured outputs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-grok-build/CHANGELOG.md b/packages/ai-grok-build/CHANGELOG.md
index bb7a7132f..8aff77133 100644
--- a/packages/ai-grok-build/CHANGELOG.md
+++ b/packages/ai-grok-build/CHANGELOG.md
@@ -1,5 +1,14 @@
 # @tanstack/ai-grok-build
 
+## 0.2.4
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-acp@0.2.4
+  - @tanstack/ai-sandbox@0.2.5
+
 ## 0.2.3
 
 ### Patch Changes
diff --git a/packages/ai-grok-build/package.json b/packages/ai-grok-build/package.json
index 7aeb00977..274841c7c 100644
--- a/packages/ai-grok-build/package.json
+++ b/packages/ai-grok-build/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-grok-build",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "description": "Grok Build harness adapter for TanStack AI — run Grok Build as a chat backend with local tool execution and stateful sessions.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-grok/CHANGELOG.md b/packages/ai-grok/CHANGELOG.md
index 80eb945ea..8164181ef 100644
--- a/packages/ai-grok/CHANGELOG.md
+++ b/packages/ai-grok/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-grok
 
+## 0.14.10
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/openai-base@0.9.10
+
 ## 0.14.9
 
 ### Patch Changes
diff --git a/packages/ai-grok/package.json b/packages/ai-grok/package.json
index caae1ee0c..3b32629da 100644
--- a/packages/ai-grok/package.json
+++ b/packages/ai-grok/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-grok",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "description": "xAI Grok adapter for TanStack AI chat, image generation, realtime, and structured outputs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-groq/CHANGELOG.md b/packages/ai-groq/CHANGELOG.md
index b3a32a8cc..90b200e11 100644
--- a/packages/ai-groq/CHANGELOG.md
+++ b/packages/ai-groq/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-groq
 
+## 0.5.4
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/openai-base@0.9.10
+
 ## 0.5.3
 
 ### Patch Changes
diff --git a/packages/ai-groq/package.json b/packages/ai-groq/package.json
index 2f1aaef65..2bd567d15 100644
--- a/packages/ai-groq/package.json
+++ b/packages/ai-groq/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-groq",
-  "version": "0.5.3",
+  "version": "0.5.4",
   "description": "Groq adapter for TanStack AI low-latency chat, tool calling, and structured outputs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-isolate-cloudflare/CHANGELOG.md b/packages/ai-isolate-cloudflare/CHANGELOG.md
index eacf35de7..f8ea64db3 100644
--- a/packages/ai-isolate-cloudflare/CHANGELOG.md
+++ b/packages/ai-isolate-cloudflare/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-isolate-cloudflare
 
+## 0.2.39
+
+### Patch Changes
+
+- Updated dependencies []:
+  - @tanstack/ai-code-mode@0.3.9
+
 ## 0.2.38
 
 ### Patch Changes
diff --git a/packages/ai-isolate-cloudflare/package.json b/packages/ai-isolate-cloudflare/package.json
index 726e3b45c..5ee9a3fca 100644
--- a/packages/ai-isolate-cloudflare/package.json
+++ b/packages/ai-isolate-cloudflare/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-isolate-cloudflare",
-  "version": "0.2.38",
+  "version": "0.2.39",
   "description": "Cloudflare Workers sandbox driver for TanStack AI Code Mode TypeScript execution at the edge.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-isolate-node/CHANGELOG.md b/packages/ai-isolate-node/CHANGELOG.md
index 355dae5c8..97bdb1d36 100644
--- a/packages/ai-isolate-node/CHANGELOG.md
+++ b/packages/ai-isolate-node/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-isolate-node
 
+## 0.1.48
+
+### Patch Changes
+
+- Updated dependencies []:
+  - @tanstack/ai-code-mode@0.3.9
+
 ## 0.1.47
 
 ### Patch Changes
diff --git a/packages/ai-isolate-node/package.json b/packages/ai-isolate-node/package.json
index 18b95cf4c..1c484f827 100644
--- a/packages/ai-isolate-node/package.json
+++ b/packages/ai-isolate-node/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-isolate-node",
-  "version": "0.1.47",
+  "version": "0.1.48",
   "description": "Node.js isolated-vm sandbox driver for TanStack AI Code Mode TypeScript execution.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-isolate-quickjs/CHANGELOG.md b/packages/ai-isolate-quickjs/CHANGELOG.md
index 21b609c72..77cf64afc 100644
--- a/packages/ai-isolate-quickjs/CHANGELOG.md
+++ b/packages/ai-isolate-quickjs/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-isolate-quickjs
 
+## 0.1.48
+
+### Patch Changes
+
+- Updated dependencies []:
+  - @tanstack/ai-code-mode@0.3.9
+
 ## 0.1.47
 
 ### Patch Changes
diff --git a/packages/ai-isolate-quickjs/package.json b/packages/ai-isolate-quickjs/package.json
index 27a7640f6..84a1e3c16 100644
--- a/packages/ai-isolate-quickjs/package.json
+++ b/packages/ai-isolate-quickjs/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-isolate-quickjs",
-  "version": "0.1.47",
+  "version": "0.1.48",
   "description": "QuickJS WASM sandbox driver for TanStack AI Code Mode TypeScript execution across runtimes.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-mcp/CHANGELOG.md b/packages/ai-mcp/CHANGELOG.md
index 68604fba8..58862b5d0 100644
--- a/packages/ai-mcp/CHANGELOG.md
+++ b/packages/ai-mcp/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-mcp
 
+## 0.2.6
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.2.5
 
 ### Patch Changes
diff --git a/packages/ai-mcp/package.json b/packages/ai-mcp/package.json
index 860f77701..13fb511e1 100644
--- a/packages/ai-mcp/package.json
+++ b/packages/ai-mcp/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-mcp",
-  "version": "0.2.5",
+  "version": "0.2.6",
   "description": "Host-side Model Context Protocol client for TanStack AI: discover and run MCP server tools, resources, and prompts in any adapter's chat() loop, with generated end-to-end types.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-memory/CHANGELOG.md b/packages/ai-memory/CHANGELOG.md
new file mode 100644
index 000000000..9e7919853
--- /dev/null
+++ b/packages/ai-memory/CHANGELOG.md
@@ -0,0 +1,88 @@
+# @tanstack/ai-memory
+
+## 0.1.0
+
+### Minor Changes
+
+- [#541](https://github.com/TanStack/ai/pull/541) [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4) - **Surface server-side memory state in the TanStack AI DevTools.**
+
+  The DevTools panel now has a **Memory** tab for any chat wired with
+  `memoryMiddleware`. It shows, per scope (session), an operations timeline (each
+  turn's recall — query, fragment count, injected system-prompt size, whether
+  memory tools were exposed, duration) and the current stored records/facts when
+  the adapter implements the optional `inspect`/`listFacts` methods.
+
+  Because memory runs on the server (whose event bus never reaches the browser),
+  the middleware transports its state to the panel over the chat stream as a
+  `memory:state` `CUSTOM` event, which `@tanstack/ai-client`'s devtools bridge
+  re-emits as browser `memory:*` events — the same pattern generation results use.
+  The snapshot reflects memory as of the start of each turn; opening the panel
+  mid-conversation replays the latest state so the tab isn't empty.
+  - `@tanstack/ai-memory` — `memoryMiddleware` injects a `memory:state` `CUSTOM`
+    chunk carrying recall metrics + an `inspect`/`listFacts` snapshot; exports
+    `MEMORY_STATE_EVENT` and `MemoryStateEventValue`.
+  - `@tanstack/ai-event-client` — adds the `memory:snapshot` devtools event.
+  - `@tanstack/ai-client` — the chat devtools bridge re-emits `memory:*` from the
+    transported chunk and replays the last snapshot on `devtools:request-state`.
+  - `@tanstack/ai-devtools-core` — new Memory tab + per-scope memory store slice.
+
+- [#541](https://github.com/TanStack/ai/pull/541) [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4) - **Add server-side memory via a `recall`/`save` adapter contract in `@tanstack/ai-memory`.**
+
+  Memory is now a single, provider-agnostic contract with two verbs — `recall` and
+  `save` — which is the shape every memory backend (in-process, Redis, and hosted
+  vendors) naturally exposes. `memoryMiddleware` recalls relevant memory into the
+  system prompt (and optionally injects vendor tools) before the model runs, then
+  defers `save` of the finished turn via `ctx.defer` so streaming is never blocked.
+  Extraction, ranking, and rendering live inside each adapter — the middleware is thin.
+
+  `@tanstack/ai-memory` (new package) — everything ships here:
+  - Root: `memoryMiddleware`, the `MemoryAdapter` contract
+    (`recall` / `save` / optional `inspect` / `listFacts`), and the `MemoryScope` /
+    `MemoryTurn` / `RecallResult` / `SaveReceipt` types.
+  - `@tanstack/ai-memory/in-memory` → `inMemory()` — zero-dependency adapter for dev,
+    tests, and single-process demos. Pass an `embedder` for semantic scoring and/or an
+    `extract` function to persist derived facts.
+  - `@tanstack/ai-memory/redis` → `redis({ redis, prefix? })` — production adapter for
+    plain Redis. `ioredis` wires in directly; `redis` (node-redis v4+) via the
+    `fromNodeRedis(client)` wrapper. Both are optional peer dependencies.
+  - `@tanstack/ai-memory/hindsight` → `hindsight()`, `@tanstack/ai-memory/mem0` →
+    `mem0()`, `@tanstack/ai-memory/honcho` → `honcho()` — hosted-vendor adapters. Their
+    SDKs (`@vectorize-io/hindsight-client`, `@honcho-ai/sdk`) are optional peers loaded
+    lazily; mem0 talks to its server over plain HTTP (no SDK). Vendors can expose LLM
+    tools through `recall` (e.g. hindsight's retain/recall/reflect).
+  - A shared `recall`/`save` contract-test suite (`@tanstack/ai-memory/tests/contract`)
+    that any adapter — including third-party ones — can run.
+
+  `@tanstack/ai`:
+  - **Removes the (unreleased) `@tanstack/ai/memory` subpath.** The middleware,
+    contract, and helpers all moved to `@tanstack/ai-memory`.
+
+  `@tanstack/ai-event-client`:
+  - The five `memory:*` devtools events (`memory:retrieve:started` / `:completed`,
+    `memory:persist:started` / `:completed`, `memory:error`) now carry recall/save
+    payloads (adapter id, fragment/receipt counts, `phase: 'recall' | 'save'`).
+
+- [#991](https://github.com/TanStack/ai/pull/991) [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371) - **Align `MemoryScope` to the shared `Scope` type (`threadId`).**
+
+  `MemoryScope` is now an alias of `Scope` from `@tanstack/ai` so memory and
+  persistence share one isolation vocabulary. The conversation key is
+  `threadId` (required); optional dims are `userId`, `tenantId`, and reserved
+  `namespace`. There is no public `sessionId` on memory scope — hard cut while
+  `@tanstack/ai-memory` is still `0.x` / unreleased.
+  - `@tanstack/ai-memory` — `export type MemoryScope = Scope`. Built-in adapters
+    (`inMemory`, `redis`) and middleware use `threadId`; `sameScope` also matches
+    `tenantId` when present on the query. Redis index keys are now
+    `{prefix}:index:{tenantId|_}:{userId|_}:{threadId}` (escaped). Hindsight banks
+    use `{user}__{threadId}`. Anyone who wrote Redis rows under the pre-rename
+    layout needs to reindex or wipe — keys are not dual-read.
+  - `@tanstack/ai-event-client` — `MemoryScopeLite` is
+    `{ threadId?, userId?, tenantId? }` (devtools telemetry; not an isolation
+    authority).
+  - `@tanstack/ai-client` / `@tanstack/ai-devtools-core` — memory event payloads
+    and the Memory panel registry follow the same `threadId` field names.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-event-client@0.7.0
diff --git a/packages/ai-memory/package.json b/packages/ai-memory/package.json
index c96b8256b..18a3b99a6 100644
--- a/packages/ai-memory/package.json
+++ b/packages/ai-memory/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-memory",
-  "version": "0.0.0",
+  "version": "0.1.0",
   "description": "Pluggable memory adapters for TanStack AI memoryMiddleware",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-mistral/CHANGELOG.md b/packages/ai-mistral/CHANGELOG.md
index 5bd877325..ee97d4a4a 100644
--- a/packages/ai-mistral/CHANGELOG.md
+++ b/packages/ai-mistral/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-mistral
 
+## 0.2.4
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.2.3
 
 ### Patch Changes
diff --git a/packages/ai-mistral/package.json b/packages/ai-mistral/package.json
index 177210044..1d5c84639 100644
--- a/packages/ai-mistral/package.json
+++ b/packages/ai-mistral/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-mistral",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "type": "module",
   "description": "Mistral adapter for TanStack AI",
   "author": "",
diff --git a/packages/ai-ollama/CHANGELOG.md b/packages/ai-ollama/CHANGELOG.md
index 80f9fa2ef..4243e86fc 100644
--- a/packages/ai-ollama/CHANGELOG.md
+++ b/packages/ai-ollama/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-ollama
 
+## 0.8.17
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.8.16
 
 ### Patch Changes
diff --git a/packages/ai-ollama/package.json b/packages/ai-ollama/package.json
index 8b414b3a4..f57636beb 100644
--- a/packages/ai-ollama/package.json
+++ b/packages/ai-ollama/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-ollama",
-  "version": "0.8.16",
+  "version": "0.8.17",
   "description": "Ollama adapter for TanStack AI local LLM chat, tool calling, and structured outputs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-openai/CHANGELOG.md b/packages/ai-openai/CHANGELOG.md
index 90c8fa673..9223985b2 100644
--- a/packages/ai-openai/CHANGELOG.md
+++ b/packages/ai-openai/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-openai
 
+## 0.17.2
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/openai-base@0.9.10
+
 ## 0.17.1
 
 ### Patch Changes
@@ -21,7 +29,9 @@
   ```ts
   createOpenaiImage('gpt-image-1', apiKey, { allowUrlFetch: true })
   createOpenaiVideo('sora-2', apiKey, { allowUrlFetch: true })
-  createGeminiVideo('veo-3.1-generate-preview', apiKey, { allowUrlFetch: true })
+  createGeminiVideo('veo-3.1-generate-preview', apiKey, {
+    allowUrlFetch: true,
+  })
   ```
 
   Migration: if you passed HTTP(S) URL image inputs to these adapters, either fetch the bytes yourself and pass a `data:` URI, pass a `gs://` reference (Veo), or set `allowUrlFetch: true`.
diff --git a/packages/ai-openai/package.json b/packages/ai-openai/package.json
index c10867be9..7ec0cc798 100644
--- a/packages/ai-openai/package.json
+++ b/packages/ai-openai/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-openai",
-  "version": "0.17.1",
+  "version": "0.17.2",
   "description": "OpenAI adapter for TanStack AI chat, tools, images, video, speech, transcription, realtime, and structured outputs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-opencode/CHANGELOG.md b/packages/ai-opencode/CHANGELOG.md
index 809af39b4..decf9b77f 100644
--- a/packages/ai-opencode/CHANGELOG.md
+++ b/packages/ai-opencode/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-opencode
 
+## 0.2.4
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-sandbox@0.2.5
+
 ## 0.2.3
 
 ### Patch Changes
diff --git a/packages/ai-opencode/package.json b/packages/ai-opencode/package.json
index 091203639..ff8583692 100644
--- a/packages/ai-opencode/package.json
+++ b/packages/ai-opencode/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-opencode",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "description": "OpenCode harness adapter for TanStack AI — run OpenCode as a chat backend with local tool execution and stateful sessions.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-openrouter/CHANGELOG.md b/packages/ai-openrouter/CHANGELOG.md
index 49d91c73f..6c0a843f5 100644
--- a/packages/ai-openrouter/CHANGELOG.md
+++ b/packages/ai-openrouter/CHANGELOG.md
@@ -1,5 +1,14 @@
 # @tanstack/ai-openrouter
 
+## 0.15.11
+
+### Patch Changes
+
+- [#888](https://github.com/TanStack/ai/pull/888) [`50d7a7d`](https://github.com/TanStack/ai/commit/50d7a7da319bd39c578548390a723302f90068c7) - Update model metadata from OpenRouter API
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.15.10
 
 ### Patch Changes
diff --git a/packages/ai-openrouter/package.json b/packages/ai-openrouter/package.json
index 5aa817ade..d3d4d8506 100644
--- a/packages/ai-openrouter/package.json
+++ b/packages/ai-openrouter/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-openrouter",
-  "version": "0.15.10",
+  "version": "0.15.11",
   "description": "TanStack AI adapter for OpenRouter chat, provider tools, structured outputs, and access to hundreds of LLMs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-preact/CHANGELOG.md b/packages/ai-preact/CHANGELOG.md
index 1f66e75ad..c937bfa90 100644
--- a/packages/ai-preact/CHANGELOG.md
+++ b/packages/ai-preact/CHANGELOG.md
@@ -1,5 +1,26 @@
 # @tanstack/ai-preact
 
+## 0.12.0
+
+### Minor Changes
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
+  client-tool execution, with typed bound resolvers, atomic batches, and
+  structured errors. Interrupts run ephemerally by resuming from the full client
+  message history in a fresh child run — no persistence required.
+
+  This changes native approval and client-tool streams from legacy custom events
+  to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
+  `pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
+  legacy event readers remain as limited compatibility surfaces for migration;
+  `addToolResult` remains supported.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-client@0.23.0
+
 ## 0.11.1
 
 ### Patch Changes
diff --git a/packages/ai-preact/package.json b/packages/ai-preact/package.json
index 3fae2bba4..4bc2b91e5 100644
--- a/packages/ai-preact/package.json
+++ b/packages/ai-preact/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-preact",
-  "version": "0.11.1",
+  "version": "0.12.0",
   "description": "Preact hooks for TanStack AI streaming chat and typed messages.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-react-ui/CHANGELOG.md b/packages/ai-react-ui/CHANGELOG.md
index 1de3edb2c..fa89c8b88 100644
--- a/packages/ai-react-ui/CHANGELOG.md
+++ b/packages/ai-react-ui/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-react-ui
 
+## 0.8.16
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288)]:
+  - @tanstack/ai-client@0.23.0
+  - @tanstack/ai-react@0.19.0
+
 ## 0.8.15
 
 ### Patch Changes
diff --git a/packages/ai-react-ui/package.json b/packages/ai-react-ui/package.json
index 05058a215..2747a47e7 100644
--- a/packages/ai-react-ui/package.json
+++ b/packages/ai-react-ui/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-react-ui",
-  "version": "0.8.15",
+  "version": "0.8.16",
   "description": "Headless React components for building TanStack AI chat interfaces with streamed message parts.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-react/CHANGELOG.md b/packages/ai-react/CHANGELOG.md
index 25bdce91d..24f4e9de4 100644
--- a/packages/ai-react/CHANGELOG.md
+++ b/packages/ai-react/CHANGELOG.md
@@ -1,5 +1,26 @@
 # @tanstack/ai-react
 
+## 0.19.0
+
+### Minor Changes
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
+  client-tool execution, with typed bound resolvers, atomic batches, and
+  structured errors. Interrupts run ephemerally by resuming from the full client
+  message history in a fresh child run — no persistence required.
+
+  This changes native approval and client-tool streams from legacy custom events
+  to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
+  `pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
+  legacy event readers remain as limited compatibility surfaces for migration;
+  `addToolResult` remains supported.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-client@0.23.0
+
 ## 0.18.1
 
 ### Patch Changes
diff --git a/packages/ai-react/package.json b/packages/ai-react/package.json
index 8cf8747e4..1940f79ff 100644
--- a/packages/ai-react/package.json
+++ b/packages/ai-react/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-react",
-  "version": "0.18.1",
+  "version": "0.19.0",
   "description": "React hooks for TanStack AI streaming chat, realtime voice, structured outputs, and media generation.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-sandbox-cloudflare/CHANGELOG.md b/packages/ai-sandbox-cloudflare/CHANGELOG.md
index 214f16495..aca727bd9 100644
--- a/packages/ai-sandbox-cloudflare/CHANGELOG.md
+++ b/packages/ai-sandbox-cloudflare/CHANGELOG.md
@@ -1,5 +1,14 @@
 # @tanstack/ai-sandbox-cloudflare
 
+## 0.2.5
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-sandbox@0.2.5
+  - @tanstack/ai-sandbox-local-process@0.2.0
+
 ## 0.2.4
 
 ### Patch Changes
diff --git a/packages/ai-sandbox-cloudflare/package.json b/packages/ai-sandbox-cloudflare/package.json
index e8c644836..f47f6f1ac 100644
--- a/packages/ai-sandbox-cloudflare/package.json
+++ b/packages/ai-sandbox-cloudflare/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-sandbox-cloudflare",
-  "version": "0.2.4",
+  "version": "0.2.5",
   "description": "Cloudflare sandbox provider for TanStack AI — run harness adapters inside Cloudflare Containers (edge) through the uniform SandboxHandle.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-sandbox/CHANGELOG.md b/packages/ai-sandbox/CHANGELOG.md
index 6f3ee0dc9..2b4be3b3d 100644
--- a/packages/ai-sandbox/CHANGELOG.md
+++ b/packages/ai-sandbox/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-sandbox
 
+## 0.2.5
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.2.4
 
 ### Patch Changes
diff --git a/packages/ai-sandbox/package.json b/packages/ai-sandbox/package.json
index 924b25279..6fab07e1b 100644
--- a/packages/ai-sandbox/package.json
+++ b/packages/ai-sandbox/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-sandbox",
-  "version": "0.2.4",
+  "version": "0.2.5",
   "description": "Provider-agnostic sandbox layer for TanStack AI — run harness adapters inside isolated sandboxes (defineSandbox, defineWorkspace, withSandbox) with a uniform SandboxHandle, workspace bootstrap, policy, and resumable lifecycle.",
   "author": "",
   "license": "MIT",
diff --git a/packages/ai-solid-ui/CHANGELOG.md b/packages/ai-solid-ui/CHANGELOG.md
index 29c42b34b..511b609e3 100644
--- a/packages/ai-solid-ui/CHANGELOG.md
+++ b/packages/ai-solid-ui/CHANGELOG.md
@@ -1,5 +1,13 @@
 # @tanstack/ai-solid-ui
 
+## 0.7.15
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288)]:
+  - @tanstack/ai-client@0.23.0
+  - @tanstack/ai-solid@0.16.0
+
 ## 0.7.14
 
 ### Patch Changes
diff --git a/packages/ai-solid-ui/package.json b/packages/ai-solid-ui/package.json
index 17cdecd3b..5670e789d 100644
--- a/packages/ai-solid-ui/package.json
+++ b/packages/ai-solid-ui/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-solid-ui",
-  "version": "0.7.14",
+  "version": "0.7.15",
   "description": "Headless Solid components for building TanStack AI chat interfaces with streamed message parts.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-solid/CHANGELOG.md b/packages/ai-solid/CHANGELOG.md
index 1e502f29a..7c560cead 100644
--- a/packages/ai-solid/CHANGELOG.md
+++ b/packages/ai-solid/CHANGELOG.md
@@ -1,5 +1,26 @@
 # @tanstack/ai-solid
 
+## 0.16.0
+
+### Minor Changes
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
+  client-tool execution, with typed bound resolvers, atomic batches, and
+  structured errors. Interrupts run ephemerally by resuming from the full client
+  message history in a fresh child run — no persistence required.
+
+  This changes native approval and client-tool streams from legacy custom events
+  to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
+  `pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
+  legacy event readers remain as limited compatibility surfaces for migration;
+  `addToolResult` remains supported.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-client@0.23.0
+
 ## 0.15.1
 
 ### Patch Changes
diff --git a/packages/ai-solid/package.json b/packages/ai-solid/package.json
index 2518689b1..f03de7491 100644
--- a/packages/ai-solid/package.json
+++ b/packages/ai-solid/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-solid",
-  "version": "0.15.1",
+  "version": "0.16.0",
   "description": "Solid hooks for TanStack AI streaming chat, structured outputs, and media generation.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-svelte/CHANGELOG.md b/packages/ai-svelte/CHANGELOG.md
index 514279b95..a2b29c1f7 100644
--- a/packages/ai-svelte/CHANGELOG.md
+++ b/packages/ai-svelte/CHANGELOG.md
@@ -1,5 +1,26 @@
 # @tanstack/ai-svelte
 
+## 0.16.0
+
+### Minor Changes
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
+  client-tool execution, with typed bound resolvers, atomic batches, and
+  structured errors. Interrupts run ephemerally by resuming from the full client
+  message history in a fresh child run — no persistence required.
+
+  This changes native approval and client-tool streams from legacy custom events
+  to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
+  `pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
+  legacy event readers remain as limited compatibility surfaces for migration;
+  `addToolResult` remains supported.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-client@0.23.0
+
 ## 0.15.1
 
 ### Patch Changes
diff --git a/packages/ai-svelte/package.json b/packages/ai-svelte/package.json
index ddb1e59b5..2c2606d6d 100644
--- a/packages/ai-svelte/package.json
+++ b/packages/ai-svelte/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-svelte",
-  "version": "0.15.1",
+  "version": "0.16.0",
   "description": "Svelte 5 bindings for TanStack AI streaming chat, structured outputs, and media generation.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-vue-ui/CHANGELOG.md b/packages/ai-vue-ui/CHANGELOG.md
index c8db6f736..b81b32a14 100644
--- a/packages/ai-vue-ui/CHANGELOG.md
+++ b/packages/ai-vue-ui/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/ai-vue-ui
 
+## 0.2.35
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a)]:
+  - @tanstack/ai-vue@0.16.0
+
 ## 0.2.34
 
 ### Patch Changes
diff --git a/packages/ai-vue-ui/package.json b/packages/ai-vue-ui/package.json
index 9036d8330..86801bccb 100644
--- a/packages/ai-vue-ui/package.json
+++ b/packages/ai-vue-ui/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-vue-ui",
-  "version": "0.2.34",
+  "version": "0.2.35",
   "description": "Headless Vue components for building TanStack AI chat interfaces with streamed message parts.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai-vue/CHANGELOG.md b/packages/ai-vue/CHANGELOG.md
index e7ed5fa20..6a2e02738 100644
--- a/packages/ai-vue/CHANGELOG.md
+++ b/packages/ai-vue/CHANGELOG.md
@@ -1,5 +1,26 @@
 # @tanstack/ai-vue
 
+## 0.16.0
+
+### Minor Changes
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
+  client-tool execution, with typed bound resolvers, atomic batches, and
+  structured errors. Interrupts run ephemerally by resuming from the full client
+  message history in a fresh child run — no persistence required.
+
+  This changes native approval and client-tool streams from legacy custom events
+  to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
+  `pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
+  legacy event readers remain as limited compatibility surfaces for migration;
+  `addToolResult` remains supported.
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+  - @tanstack/ai-client@0.23.0
+
 ## 0.15.1
 
 ### Patch Changes
diff --git a/packages/ai-vue/package.json b/packages/ai-vue/package.json
index fc57375f0..abcce741b 100644
--- a/packages/ai-vue/package.json
+++ b/packages/ai-vue/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai-vue",
-  "version": "0.15.1",
+  "version": "0.16.0",
   "description": "Vue composables for TanStack AI streaming chat, structured outputs, and media generation.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/ai/CHANGELOG.md b/packages/ai/CHANGELOG.md
index a95900586..c5cedc767 100644
--- a/packages/ai/CHANGELOG.md
+++ b/packages/ai/CHANGELOG.md
@@ -1,5 +1,177 @@
 # @tanstack/ai
 
+## 0.43.0
+
+### Minor Changes
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Adopt the AG-UI interrupt lifecycle for tool approvals, generic responses, and
+  client-tool execution, with typed bound resolvers, atomic batches, and
+  structured errors. Interrupts run ephemerally by resuming from the full client
+  message history in a fresh child run — no persistence required.
+
+  This changes native approval and client-tool streams from legacy custom events
+  to snapshot-plus-`RUN_FINISHED` interrupt outcomes. Deprecated
+  `pendingInterrupts`, `addToolApprovalResponse`, raw `resumeInterrupts`, and
+  legacy event readers remain as limited compatibility surfaces for migration;
+  `addToolResult` remains supported.
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Make interrupt ownership explicit rather than assumed.
+
+  An AG-UI `Interrupt` is a shared envelope — a workflow engine's durable
+  approval or another agent framework's pause can arrive on the same stream. What
+  makes a pause resumable through `chat()` is the binding this package attaches
+  under `tanstack:interruptBinding`.
+  - Interrupts that carry no binding this client understands now surface as
+    `kind: 'unbound'` with `canResolve: false`, instead of being given a
+    synthesized binding and rendered as resolvable generic interrupts. Resolving
+    those produced an answer submitted against a run with nothing pending, which
+    failed as `unknown-interrupt` only after the user had filled in the form.
+    Unbound items never block submission of the interrupts that are yours.
+  - The binding carries a wire version (`INTERRUPT_BINDING_VERSION`). Readers
+    reject a version they don't recognise rather than duck-typing its fields. A
+    binding written before the field existed is still read.
+  - `INTERRUPT_BINDING_METADATA_KEY`, `withInterruptBinding()` and
+    `readInterruptBinding()` are exported, so anything producing an interrupt this
+    package must later resume attaches the binding through a supported API
+    instead of copying the metadata key.
+  - Interrupt classification is driven by the binding alone. `Interrupt.reason` is
+    free-form AG-UI text another producer can also use, so it is now a display
+    hint only and never decides ownership.
+  - The interrupt protocol surface is enumerated instead of `export *`. The
+    unimplemented durable-recovery contract (`InterruptRecoveryStateV1`,
+    `InterruptRecoveryQuery`, the never-called `loadInterruptState` adapter hook,
+    and the `persistence-required` / `atomic-commit-unsupported` /
+    `recovery-unavailable` error codes) is removed rather than published.
+
+- [#970](https://github.com/TanStack/ai/pull/970) [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a) - Interrupts: the application owns wire-schema validation, and the hashing
+  dependency is gone.
+
+  The library no longer transforms a generic interrupt's wire JSON Schema into a
+  validator or validates the resolved value against it, on either the client or
+  the server. Whatever you pass to `resolveInterrupt` (client) or send in the
+  `resume` batch (server) flows through as-is. Validate it yourself if you need to
+  trust it, e.g. with `z.fromJSONSchema(interrupt.responseSchema).safeParse(value)`
+  on the client and your own check on the server. Validation of a tool's
+  code-authored Standard Schema (`approvalSchema` / `inputSchema`) is unchanged.
+
+  This drops the `ajv` and `ajv-formats` dependencies. Interrupt binding hashes and
+  resolution fingerprints now use a small bundled SHA-256 instead of
+  `@noble/hashes`, so that dependency is gone too. The wire hash shape
+  (`sha256:<hex>`) is unchanged.
+
+- [#541](https://github.com/TanStack/ai/pull/541) [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4) - **Add server-side memory via a `recall`/`save` adapter contract in `@tanstack/ai-memory`.**
+
+  Memory is now a single, provider-agnostic contract with two verbs — `recall` and
+  `save` — which is the shape every memory backend (in-process, Redis, and hosted
+  vendors) naturally exposes. `memoryMiddleware` recalls relevant memory into the
+  system prompt (and optionally injects vendor tools) before the model runs, then
+  defers `save` of the finished turn via `ctx.defer` so streaming is never blocked.
+  Extraction, ranking, and rendering live inside each adapter — the middleware is thin.
+
+  `@tanstack/ai-memory` (new package) — everything ships here:
+  - Root: `memoryMiddleware`, the `MemoryAdapter` contract
+    (`recall` / `save` / optional `inspect` / `listFacts`), and the `MemoryScope` /
+    `MemoryTurn` / `RecallResult` / `SaveReceipt` types.
+  - `@tanstack/ai-memory/in-memory` → `inMemory()` — zero-dependency adapter for dev,
+    tests, and single-process demos. Pass an `embedder` for semantic scoring and/or an
+    `extract` function to persist derived facts.
+  - `@tanstack/ai-memory/redis` → `redis({ redis, prefix? })` — production adapter for
+    plain Redis. `ioredis` wires in directly; `redis` (node-redis v4+) via the
+    `fromNodeRedis(client)` wrapper. Both are optional peer dependencies.
+  - `@tanstack/ai-memory/hindsight` → `hindsight()`, `@tanstack/ai-memory/mem0` →
+    `mem0()`, `@tanstack/ai-memory/honcho` → `honcho()` — hosted-vendor adapters. Their
+    SDKs (`@vectorize-io/hindsight-client`, `@honcho-ai/sdk`) are optional peers loaded
+    lazily; mem0 talks to its server over plain HTTP (no SDK). Vendors can expose LLM
+    tools through `recall` (e.g. hindsight's retain/recall/reflect).
+  - A shared `recall`/`save` contract-test suite (`@tanstack/ai-memory/tests/contract`)
+    that any adapter — including third-party ones — can run.
+
+  `@tanstack/ai`:
+  - **Removes the (unreleased) `@tanstack/ai/memory` subpath.** The middleware,
+    contract, and helpers all moved to `@tanstack/ai-memory`.
+
+  `@tanstack/ai-event-client`:
+  - The five `memory:*` devtools events (`memory:retrieve:started` / `:completed`,
+    `memory:persist:started` / `:completed`, `memory:error`) now carry recall/save
+    payloads (adapter id, fragment/receipt counts, `phase: 'recall' | 'save'`).
+
+- [#955](https://github.com/TanStack/ai/pull/955) [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288) - Resumable streams: reconnect to an in-flight SSE **or NDJSON** response without
+  re-running the provider.
+
+  `toServerSentEventsResponse` and `toHttpResponse` both accept a
+  `durability: { adapter, batch }` option. The adapter (`StreamDurability`)
+  records every chunk to an ordered log before delivery and tags each event with
+  an opaque, adapter-owned offset — an SSE `id:` line, or the `id` of an NDJSON
+  `{ id, chunk }` envelope (NDJSON has no native event-id). A reconnect
+  (`Last-Event-ID`) or an explicit `?offset` read replays strictly after that
+  offset from the log — the lazy provider stream is never iterated on resume.
+  Producers terminalize the log on cancellation and failure (`RUN_ERROR` append
+  - `close()`) and on completion when the source stream emits its own terminal
+    event (`chat()` always does), so readers are never parked on a dead run.
+
+  Two adapters ship: `memoryStream(request)` in `@tanstack/ai` (process-local,
+  for development and tests) and the new `@tanstack/ai-durable-stream` package,
+  a Durable Streams protocol adapter for production backends.
+
+  For the `GET` handler that a reload or a second tab reconnects to,
+  `resumeServerSentEventsResponse({ adapter })` and `resumeHttpResponse({ adapter })`
+  replay a run straight from the durability log. They need no producer stream and
+  return a 400 when the request carries no resume offset.
+
+  On the client, all four HTTP adapters are now resumable — `fetchServerSentEvents`,
+  `fetchHttpStream`, `xhrServerSentEvents`, and `xhrHttpStream`. Each tracks the
+  per-event offset, auto-reconnects with `Last-Event-ID`, de-duplicates the
+  replayed prefix, and exposes `joinRun(runId)` to attach to an in-flight or
+  finished run from the start (read-only GET with `offset=-1`). Untagged streams
+  behave exactly as before. A durable run that ends with no terminal event and no
+  forward progress now throws `DurableStreamIncompleteError` instead of hanging.
+
+  Reconnection and durability are bounded so failures surface rather than hang or
+  loop:
+  - `memoryStream` evicts completed logs after a grace window (unbounded growth
+    is gone); resuming an expired/unknown run throws, and a from-start join to a
+    run that never produces fails after `MemoryStreamOptions.firstChunkDeadlineMs`.
+  - all four HTTP adapters accept `reconnect: { maxAttempts, delayMs }` — a
+    throttle plus a ceiling on CONSECUTIVE no-progress reconnects (default 5;
+    forward progress resets it) that fails with the new `StreamReconnectLimitError`
+    instead of reconnecting endlessly, without penalizing a healthy long-lived run.
+  - `durableStream` accepts `reconnect: { maxReadFailures, delayMs }` to bound its
+    read-retry loop, and `server` is now optional when `fetch` is provided (e.g. a
+    Cloudflare service binding).
+  - `toServerSentEventsResponse` accepts `debug` to record durability terminal /
+    close failures server-side, where a replaying joiner cannot observe them.
+
+- [#980](https://github.com/TanStack/ai/pull/980) [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa) - **Add a shared `Scope` identity type to `@tanstack/ai`.**
+
+  `Scope` is the single identity/isolation vocabulary for the subsystems that
+  persist or recall per-conversation data — `@tanstack/ai-persistence` and
+  `@tanstack/ai-memory`. Rather than each subsystem inventing its own notion of
+  "whose data is this?", both import one type:
+
+  ```ts
+  interface Scope {
+    threadId: string // required — the single conversation key (same as ctx.threadId)
+    userId?: string // durable end-user identity; required in practice for multi-user apps
+    tenantId?: string // multi-tenant boundary
+    namespace?: string // reserved logical partition; no subsystem keys on it yet
+  }
+  ```
+
+  `threadId` is the one conversation key across the codebase (matching
+  `ChatMiddlewareContext.threadId`, with `conversationId` already deprecated in
+  favor of it) — subsystems must not introduce a second name (`sessionId`, …) for
+  the same concept. Every field is an isolation boundary and must be derived
+  server-side from trusted session state, never from client input.
+
+  Introduced ahead of the persistence and memory packages so both share one settled
+  identity contract. `@tanstack/ai-memory` now aliases `MemoryScope` to `Scope`
+  (see the memory-scope-threadid changeset).
+
+### Patch Changes
+
+- Updated dependencies [[`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371)]:
+  - @tanstack/ai-event-client@0.7.0
+
 ## 0.42.0
 
 ### Minor Changes
diff --git a/packages/ai/package.json b/packages/ai/package.json
index 31cf76284..c29a16540 100644
--- a/packages/ai/package.json
+++ b/packages/ai/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/ai",
-  "version": "0.42.0",
+  "version": "0.43.0",
   "description": "Type-safe TypeScript AI SDK for streaming chat, tool calling, agents, structured outputs, and multimodal generation.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/openai-base/CHANGELOG.md b/packages/openai-base/CHANGELOG.md
index 16a2b786a..b2139d83a 100644
--- a/packages/openai-base/CHANGELOG.md
+++ b/packages/openai-base/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/openai-base
 
+## 0.9.10
+
+### Patch Changes
+
+- Updated dependencies [[`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`3301398`](https://github.com/TanStack/ai/commit/330139878958fc5c5c167a69347c884fa35b792a), [`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`7c7aa09`](https://github.com/TanStack/ai/commit/7c7aa09a7402b45e6285ebc78a606131aec3e288), [`4ce7600`](https://github.com/TanStack/ai/commit/4ce7600d5b543d4b7e3bd6d63cdf5ecf91cdeeaa)]:
+  - @tanstack/ai@0.43.0
+
 ## 0.9.9
 
 ### Patch Changes
diff --git a/packages/openai-base/package.json b/packages/openai-base/package.json
index 63532ef4e..eb9241884 100644
--- a/packages/openai-base/package.json
+++ b/packages/openai-base/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/openai-base",
-  "version": "0.9.9",
+  "version": "0.9.10",
   "description": "Shared OpenAI SDK base adapters for TanStack AI providers using Chat Completions and Responses APIs.",
   "author": "Tanner Linsley",
   "license": "MIT",
diff --git a/packages/preact-ai-devtools/CHANGELOG.md b/packages/preact-ai-devtools/CHANGELOG.md
index adf1ffa56..fa87087e4 100644
--- a/packages/preact-ai-devtools/CHANGELOG.md
+++ b/packages/preact-ai-devtools/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/preact-ai-devtools
 
+## 0.1.68
+
+### Patch Changes
+
+- Updated dependencies [[`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371)]:
+  - @tanstack/ai-devtools-core@0.5.0
+
 ## 0.1.67
 
 ### Patch Changes
diff --git a/packages/preact-ai-devtools/package.json b/packages/preact-ai-devtools/package.json
index c33f8246f..dbe39533d 100644
--- a/packages/preact-ai-devtools/package.json
+++ b/packages/preact-ai-devtools/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/preact-ai-devtools",
-  "version": "0.1.67",
+  "version": "0.1.68",
   "description": "Preact Devtools plugin for inspecting TanStack AI chat messages, tool calls, streams, and errors.",
   "author": "tannerlinsley",
   "license": "MIT",
diff --git a/packages/react-ai-devtools/CHANGELOG.md b/packages/react-ai-devtools/CHANGELOG.md
index 7f137cae6..a150d0bb9 100644
--- a/packages/react-ai-devtools/CHANGELOG.md
+++ b/packages/react-ai-devtools/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/react-ai-devtools
 
+## 0.2.68
+
+### Patch Changes
+
+- Updated dependencies [[`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371)]:
+  - @tanstack/ai-devtools-core@0.5.0
+
 ## 0.2.67
 
 ### Patch Changes
diff --git a/packages/react-ai-devtools/package.json b/packages/react-ai-devtools/package.json
index 18dc96f51..fabdc3642 100644
--- a/packages/react-ai-devtools/package.json
+++ b/packages/react-ai-devtools/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/react-ai-devtools",
-  "version": "0.2.67",
+  "version": "0.2.68",
   "description": "React Devtools plugin for inspecting TanStack AI chat messages, tool calls, streams, and errors.",
   "author": "tannerlinsley",
   "license": "MIT",
diff --git a/packages/solid-ai-devtools/CHANGELOG.md b/packages/solid-ai-devtools/CHANGELOG.md
index b0b43c73c..6f34f18f6 100644
--- a/packages/solid-ai-devtools/CHANGELOG.md
+++ b/packages/solid-ai-devtools/CHANGELOG.md
@@ -1,5 +1,12 @@
 # @tanstack/solid-ai-devtools
 
+## 0.2.68
+
+### Patch Changes
+
+- Updated dependencies [[`347b61b`](https://github.com/TanStack/ai/commit/347b61bc788bb816bbd12287c1a426ca7def00f4), [`cc88874`](https://github.com/TanStack/ai/commit/cc88874ecb0639daa1f8a8c32be5dcc9b2749371)]:
+  - @tanstack/ai-devtools-core@0.5.0
+
 ## 0.2.67
 
 ### Patch Changes
diff --git a/packages/solid-ai-devtools/package.json b/packages/solid-ai-devtools/package.json
index 492ed6a9a..85cf57124 100644
--- a/packages/solid-ai-devtools/package.json
+++ b/packages/solid-ai-devtools/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@tanstack/solid-ai-devtools",
-  "version": "0.2.67",
+  "version": "0.2.68",
   "description": "Solid Devtools plugin for inspecting TanStack AI chat messages, tool calls, streams, and errors.",
   "author": "Tanner Linsley",
   "license": "MIT",