chore: prepare 0.8.23 release

CHANGELOG.md

@@ -1,4 +1,4 @@
-## Unreleased
+## 0.8.23
 
 * Fail Apple builds before native symbol lookup when the resolved llama.cpp
   companion does not match the core native runtime, with an actionable upgrade
@@ -9,13 +9,17 @@
   the `llamadart_llama_cpp_flutter` Apple SwiftPM checksum, and
   aligned current README/website native override docs. Updated multimodal calls
   for the matching ABI. Saved native sessions from older runtimes must be
-  regenerated. Prepared Apple companion `0.0.18` for the matching native runtime.
+  regenerated. Apple companion `0.0.18` supplies the matching native runtime.
 
 * Aligned the default WebGPU bridge assets to `v0.1.43` for Web/native
   llama.cpp `v0.4.0@5266f24da75dc449bd56cbed7addb9c8e4a6a73e` parity.
   Web `@litert-lm/core@0.15.0` and native LiteRT pins are unchanged. Immutable
   manifest: `111eefc3588842cebfe665b363378edca34924764610263e1eda5280dfcfaa27`.
 
+* Known upstream limitation: llama.cpp v0.4.0 can reject large grammar
+  repetitions, such as `root ::= "a"{2000}`. The post-v0.4.0 upstream
+  correction is tracked in `llamadart-native#76` and is not part of this release.
+
 ## 0.8.22
 
 * Updated `llamadart_llama_cpp_flutter` to `0.0.17` with the

README.md

@@ -55,18 +55,14 @@ For Dart or Flutter apps:
 
 ```yaml
 dependencies:
-  llamadart: ^0.8.22
+  llamadart: ^0.8.23
 ```
 
 Flutter iOS/macOS apps that should link Apple XCFrameworks through Swift
 Package Manager should also add the runtime companion packages they need:
 
-**Unreleased coordinated upgrade:** companion `0.0.18` requires the matching
-llama.cpp v0.4.0 core bindings; it is not compatible with published core
-`0.8.22`. For published packages, keep core `0.8.22` with companion `0.0.17`.
-The development example below requires both path overrides to the same checkout.
-Publish companion `0.0.18` only with the matching next core release, and replace
-these temporary overrides/version constraints during that coordinated release.
+Pair companion `0.0.18` with core `0.8.23` for matching llama.cpp v0.4.0
+bindings. Keep core `0.8.22` paired with companion `0.0.17`.
 
 Apple builds verify the resolved companion's SwiftPM runtime pin before native
 symbol lookup. Incompatible companions or unverified local `Artifacts`
@@ -75,15 +71,9 @@ overrides fail the build; resolve the matching companion and rerun
 
 ```yaml
 dependencies:
-  llamadart: ^0.8.22
+  llamadart: ^0.8.23
   llamadart_llama_cpp_flutter: ^0.0.18 # GGUF / llama.cpp
   llamadart_litert_lm_flutter: ^0.0.10 # Apple .litertlm / LiteRT-LM targets
-
-dependency_overrides:
-  llamadart:
-    path: /path/to/llamadart
-  llamadart_llama_cpp_flutter:
-    path: /path/to/llamadart/packages/llamadart_llama_cpp_flutter
 ```
 
 The LiteRT-LM companion manifest includes the complete iOS SwiftPM runtime

example/chat_app/pubspec.lock

@@ -413,7 +413,7 @@ packages:
       path: "../.."
       relative: true
     source: path
-    version: "0.8.22"
+    version: "0.8.23"
   llamadart_litert_lm_flutter:
     dependency: "direct main"
     description:

packages/llamadart_litert_lm_flutter/README.md

@@ -15,7 +15,7 @@ also request an x86_64 Simulator slice must exclude x86_64 for LiteRT-LM builds.
 
 ```yaml
 dependencies:
-  llamadart: ^0.8.22
+  llamadart: ^0.8.23
   llamadart_litert_lm_flutter: ^0.0.10
 ```
 

packages/llamadart_llama_cpp_flutter/README.md

@@ -7,23 +7,13 @@ Add this package to a Flutter iOS/macOS app when the app should link the
 prebuilt llama.cpp Apple XCFramework through SwiftPM instead of relying on the
 core package's native-assets fallback.
 
-**Unreleased coordinated upgrade:** companion `0.0.18` requires the matching
-llama.cpp v0.4.0 core bindings; it is not compatible with published core
-`0.8.22`. For published packages, keep core `0.8.22` with companion `0.0.17`.
-The development example below requires both path overrides to the same checkout.
-Publish companion `0.0.18` only with the matching next core release, and replace
-these temporary overrides/version constraints during that coordinated release.
+Pair companion `0.0.18` with core `0.8.23` for matching llama.cpp v0.4.0
+bindings. Keep core `0.8.22` paired with companion `0.0.17`.
 
 ```yaml
 dependencies:
-  llamadart: ^0.8.22
+  llamadart: ^0.8.23
   llamadart_llama_cpp_flutter: ^0.0.18
-
-dependency_overrides:
-  llamadart:
-    path: /path/to/llamadart
-  llamadart_llama_cpp_flutter:
-    path: /path/to/llamadart/packages/llamadart_llama_cpp_flutter
 ```
 
 This package has no runtime Dart API of its own. Import `package:llamadart`

pubspec.yaml

@@ -1,6 +1,6 @@
 name: llamadart
 description: Dart and Flutter local LLM inference with llama.cpp GGUF and LiteRT-LM across native platforms and web.
-version: 0.8.22
+version: 0.8.23
 homepage: https://github.com/leehack/llamadart
 repository: https://github.com/leehack/llamadart
 issue_tracker: https://github.com/leehack/llamadart/issues

website/docs/changelog/recent-releases.md

@@ -7,21 +7,26 @@ For canonical full release notes, use:
 
 - [`CHANGELOG.md`](https://github.com/leehack/llamadart/blob/main/CHANGELOG.md)
 
-## Unreleased
+## 0.8.23
 
 - Fail Apple builds before native symbol lookup when the resolved llama.cpp
   companion does not match the core native runtime, with actionable upgrade
   guidance.
 
 - Adopted native llama.cpp v0.4.0 with matching bindings and multimodal calls.
-  Saved native sessions from older runtimes must be regenerated. Prepared Apple
-  companion `0.0.18` for the matching native runtime.
+  Saved native sessions from older runtimes must be regenerated. Apple companion
+  `0.0.18` supplies the matching native runtime.
 
 - Aligned default WebGPU bridge assets to `v0.1.43` for Web/native
   llama.cpp `v0.4.0@5266f24da75dc449bd56cbed7addb9c8e4a6a73e` parity.
   Web `@litert-lm/core@0.15.0` and native LiteRT pins are unchanged. Immutable
   manifest: `111eefc3588842cebfe665b363378edca34924764610263e1eda5280dfcfaa27`.
 
+- Known upstream limitation: llama.cpp v0.4.0 can reject large grammar
+  repetitions, such as `root ::= "a"{2000}`. The post-v0.4.0 correction is
+  tracked in [native #76](https://github.com/leehack/llamadart-native/issues/76)
+  and is not included in this release.
+
 ## 0.8.22
 
 - Updated `llamadart_llama_cpp_flutter` to `0.0.17` with the

website/docs/getting-started/installation.md

@@ -27,18 +27,14 @@ In Xcode, set `IPHONEOS_DEPLOYMENT_TARGET = 16.4` or
 
 ```yaml
 dependencies:
-  llamadart: ^0.8.22
+  llamadart: ^0.8.23
 ```
 
 For Flutter iOS/macOS apps that should link Apple XCFrameworks through Swift
 Package Manager, also add the runtime companion packages you need:
 
-**Unreleased coordinated upgrade:** companion `0.0.18` requires the matching
-llama.cpp v0.4.0 core bindings; it is not compatible with published core
-`0.8.22`. For published packages, keep core `0.8.22` with companion `0.0.17`.
-The development example below requires both path overrides to the same checkout.
-Publish companion `0.0.18` only with the matching next core release, and replace
-these temporary overrides/version constraints during that coordinated release.
+Pair companion `0.0.18` with core `0.8.23` for matching llama.cpp v0.4.0
+bindings. Keep core `0.8.22` paired with companion `0.0.17`.
 
 Apple builds verify the resolved companion's SwiftPM runtime pin before native
 symbol lookup. Incompatible companions or unverified local `Artifacts`
@@ -47,15 +43,9 @@ overrides fail the build; resolve the matching companion and rerun
 
 ```yaml
 dependencies:
-  llamadart: ^0.8.22
+  llamadart: ^0.8.23
   llamadart_llama_cpp_flutter: ^0.0.18 # GGUF / llama.cpp
   llamadart_litert_lm_flutter: ^0.0.10 # Apple .litertlm / LiteRT-LM targets
-
-dependency_overrides:
-  llamadart:
-    path: /path/to/llamadart
-  llamadart_llama_cpp_flutter:
-    path: /path/to/llamadart/packages/llamadart_llama_cpp_flutter
 ```
 
 The companion packages are published independently from the `packages/`