chore: release main

.release-please-manifest.json

@@ -1,5 +1,6 @@
 {
-  "packages/sdk/server-ai": "0.16.1",
-  "packages/ai-providers/server-ai-langchain": "0.3.2",
-  "packages/ai-providers/server-ai-openai": "0.2.1"
+  "packages/sdk/server-ai": "0.17.0",
+  "packages/ai-providers/server-ai-langchain": "0.4.0",
+  "packages/ai-providers/server-ai-openai": "0.3.0",
+  "packages/optimization": "0.1.0"
 }

packages/ai-providers/server-ai-langchain/CHANGELOG.md

@@ -1,5 +1,40 @@
 # Changelog
 
+## [0.4.0](https://github.com/launchdarkly/python-server-sdk-ai/compare/launchdarkly-server-sdk-ai-langchain-0.3.2...launchdarkly-server-sdk-ai-langchain-0.4.0) (2026-03-26)
+
+
+### ⚠ BREAKING CHANGES
+
+* Split track_metrics_of into sync and async variants ([#112](https://github.com/launchdarkly/python-server-sdk-ai/issues/112))
+* Remove node-scoped methods from AIGraphTracker (track_node_invocation, track_tool_call, track_node_judge_response), use related AIConfigTracker methods instead
+* Extract shared utilities to openai_helper
+* Extract shared utilities to langchain_helper
+* Restructure provider factory and support additional create methods ([#102](https://github.com/launchdarkly/python-server-sdk-ai/issues/102))
+
+### Features
+
+* Add get_ai_usage_from_response to openai_helper ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Add get_tool_calls_from_response and sum_token_usage_from_messages to langchain_helper ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Add ModelRunner ABC with invoke_model() and ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* add optimization package stub ([872e81e](https://github.com/launchdarkly/python-server-sdk-ai/commit/872e81e29854ec03c434a32a287e9c94feb0b449))
+* Add optional graph_key to all LDAIConfigTracker track_* methods for graph correlation ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Add track_tool_call/track_tool_calls to LDAIConfigTracker ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Adds optimization package stub ([58b7731](https://github.com/launchdarkly/python-server-sdk-ai/commit/58b7731aa4f0efbd42ff0b93760eb357cdfe219f))
+* Deprecated Chat object in favor of ManagedModel ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Deprecated create_chat(), use create_model() on the LDAIClient ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Extract shared utilities to langchain_helper ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Extract shared utilities to openai_helper ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Introduce ManagedModel and ModelRunner (PR-3) ([#104](https://github.com/launchdarkly/python-server-sdk-ai/issues/104)) ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Restructure provider factory and support additional create methods ([#102](https://github.com/launchdarkly/python-server-sdk-ai/issues/102)) ([e6e4907](https://github.com/launchdarkly/python-server-sdk-ai/commit/e6e49076b3b89b9bf4996d3ba8e4b4c9fb7b2078))
+* Split track_metrics_of into sync and async variants ([#112](https://github.com/launchdarkly/python-server-sdk-ai/issues/112)) ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+
+
+### Bug Fixes
+
+* make AIGraphTracker.track_total_tokens accept Optional[TokenUsage], skip when None or total <= 0 ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Remove node-scoped methods from AIGraphTracker (track_node_invocation, track_tool_call, track_node_judge_response), use related AIConfigTracker methods instead ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* use time.perf_counter_ns() for sub-millisecond precision in duration calculations ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+
 ## [0.3.2](https://github.com/launchdarkly/python-server-sdk-ai/compare/launchdarkly-server-sdk-ai-langchain-0.3.1...launchdarkly-server-sdk-ai-langchain-0.3.2) (2026-03-16)
 
 

packages/ai-providers/server-ai-langchain/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "launchdarkly-server-sdk-ai-langchain"
-version = "0.3.2"
+version = "0.4.0"
 description = "LaunchDarkly AI SDK LangChain Provider"
 authors = [{name = "LaunchDarkly", email = "dev@launchdarkly.com"}]
 license = {text = "Apache-2.0"}

packages/ai-providers/server-ai-openai/CHANGELOG.md

@@ -1,5 +1,26 @@
 # Changelog
 
+## [0.3.0](https://github.com/launchdarkly/python-server-sdk-ai/compare/launchdarkly-server-sdk-ai-openai-0.2.1...launchdarkly-server-sdk-ai-openai-0.3.0) (2026-03-26)
+
+
+### ⚠ BREAKING CHANGES
+
+* Extract shared utilities to openai_helper
+* Extract shared utilities to langchain_helper
+* Restructure provider factory and support additional create methods ([#102](https://github.com/launchdarkly/python-server-sdk-ai/issues/102))
+
+### Features
+
+* Add ModelRunner ABC with invoke_model() and ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* add optimization package stub ([872e81e](https://github.com/launchdarkly/python-server-sdk-ai/commit/872e81e29854ec03c434a32a287e9c94feb0b449))
+* Adds optimization package stub ([58b7731](https://github.com/launchdarkly/python-server-sdk-ai/commit/58b7731aa4f0efbd42ff0b93760eb357cdfe219f))
+* Deprecated Chat object in favor of ManagedModel ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Deprecated create_chat(), use create_model() on the LDAIClient ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Extract shared utilities to langchain_helper ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Extract shared utilities to openai_helper ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Introduce ManagedModel and ModelRunner (PR-3) ([#104](https://github.com/launchdarkly/python-server-sdk-ai/issues/104)) ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Restructure provider factory and support additional create methods ([#102](https://github.com/launchdarkly/python-server-sdk-ai/issues/102)) ([e6e4907](https://github.com/launchdarkly/python-server-sdk-ai/commit/e6e49076b3b89b9bf4996d3ba8e4b4c9fb7b2078))
+
 ## [0.2.1](https://github.com/launchdarkly/python-server-sdk-ai/compare/launchdarkly-server-sdk-ai-openai-0.2.0...launchdarkly-server-sdk-ai-openai-0.2.1) (2026-03-16)
 
 

packages/ai-providers/server-ai-openai/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "launchdarkly-server-sdk-ai-openai"
-version = "0.2.1"
+version = "0.3.0"
 description = "LaunchDarkly AI SDK OpenAI Provider"
 authors = [{name = "LaunchDarkly", email = "dev@launchdarkly.com"}]
 license = {text = "Apache-2.0"}

packages/optimization/CHANGELOG.md

@@ -1,5 +1,19 @@
 # Changelog
 
+## 0.1.0 (2026-03-26)
+
+
+### Features
+
+* Add optimization package stub ([#109](https://github.com/launchdarkly/python-server-sdk-ai/issues/109)) ([ebd5166](https://github.com/launchdarkly/python-server-sdk-ai/commit/ebd5166d86c2d58e4c2fcc0b3fcc983eb49574e6))
+* Adds optimization package stub ([58b7731](https://github.com/launchdarkly/python-server-sdk-ai/commit/58b7731aa4f0efbd42ff0b93760eb357cdfe219f))
+* Adds optimization package stub ([cc85a05](https://github.com/launchdarkly/python-server-sdk-ai/commit/cc85a05f0e81acdc33437b51238d67ddf8a92b80))
+
+
+### Bug Fixes
+
+* consistency with other makefiles ([b9a5601](https://github.com/launchdarkly/python-server-sdk-ai/commit/b9a560110b9ef1746b1b1cff2b50ea8b90297acd))
+
 ## 0.0.0 (2026-03-24)
 
 ### Features

packages/optimization/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "launchdarkly-server-sdk-ai-optimization"
-version = "0.0.0"  # x-release-please-version
+version = "0.1.0"  # x-release-please-version
 description = "LaunchDarkly AI SDK optimization helpers"
 authors = [{name = "LaunchDarkly", email = "dev@launchdarkly.com"}]
 license = {text = "Apache-2.0"}

packages/sdk/server-ai/CHANGELOG.md

@@ -2,6 +2,41 @@
 
 All notable changes to the LaunchDarkly Python AI package will be documented in this file. This project adheres to [Semantic Versioning](http://semver.org).
 
+## [0.17.0](https://github.com/launchdarkly/python-server-sdk-ai/compare/launchdarkly-server-sdk-ai-0.16.1...launchdarkly-server-sdk-ai-0.17.0) (2026-03-26)
+
+
+### ⚠ BREAKING CHANGES
+
+* Split track_metrics_of into sync and async variants ([#112](https://github.com/launchdarkly/python-server-sdk-ai/issues/112))
+* Remove node-scoped methods from AIGraphTracker (track_node_invocation, track_tool_call, track_node_judge_response), use related AIConfigTracker methods instead
+* Extract shared utilities to openai_helper
+* Extract shared utilities to langchain_helper
+* Restructure provider factory and support additional create methods ([#102](https://github.com/launchdarkly/python-server-sdk-ai/issues/102))
+
+### Features
+
+* Add get_ai_usage_from_response to openai_helper ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Add get_tool_calls_from_response and sum_token_usage_from_messages to langchain_helper ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Add ModelRunner ABC with invoke_model() and ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* add optimization package stub ([872e81e](https://github.com/launchdarkly/python-server-sdk-ai/commit/872e81e29854ec03c434a32a287e9c94feb0b449))
+* Add optional graph_key to all LDAIConfigTracker track_* methods for graph correlation ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Add track_tool_call/track_tool_calls to LDAIConfigTracker ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Adds optimization package stub ([58b7731](https://github.com/launchdarkly/python-server-sdk-ai/commit/58b7731aa4f0efbd42ff0b93760eb357cdfe219f))
+* Deprecated Chat object in favor of ManagedModel ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Deprecated create_chat(), use create_model() on the LDAIClient ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Extract shared utilities to langchain_helper ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Extract shared utilities to openai_helper ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Introduce ManagedModel and ModelRunner (PR-3) ([#104](https://github.com/launchdarkly/python-server-sdk-ai/issues/104)) ([453c71c](https://github.com/launchdarkly/python-server-sdk-ai/commit/453c71c84adcc6b8a3e316a98907dcb511bc9d41))
+* Restructure provider factory and support additional create methods ([#102](https://github.com/launchdarkly/python-server-sdk-ai/issues/102)) ([e6e4907](https://github.com/launchdarkly/python-server-sdk-ai/commit/e6e49076b3b89b9bf4996d3ba8e4b4c9fb7b2078))
+* Split track_metrics_of into sync and async variants ([#112](https://github.com/launchdarkly/python-server-sdk-ai/issues/112)) ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+
+
+### Bug Fixes
+
+* make AIGraphTracker.track_total_tokens accept Optional[TokenUsage], skip when None or total <= 0 ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* Remove node-scoped methods from AIGraphTracker (track_node_invocation, track_tool_call, track_node_judge_response), use related AIConfigTracker methods instead ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+* use time.perf_counter_ns() for sub-millisecond precision in duration calculations ([4fab18f](https://github.com/launchdarkly/python-server-sdk-ai/commit/4fab18fa62375b6c97cb12a89225805c81ca4ee8))
+
 ## [0.16.1](https://github.com/launchdarkly/python-server-sdk-ai/compare/launchdarkly-server-sdk-ai-0.16.0...launchdarkly-server-sdk-ai-0.16.1) (2026-03-16)
 
 

packages/sdk/server-ai/PROVENANCE.md

@@ -10,7 +10,7 @@ To verify SLSA provenance attestations, we recommend using [slsa-verifier](https
 
 ```
 # Set the version of the library to verify
-VERSION=0.16.1
+VERSION=0.17.0
 ```
 
 <!-- x-release-please-end -->

packages/sdk/server-ai/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "launchdarkly-server-sdk-ai"
-version = "0.16.1"  # x-release-please-version
+version = "0.17.0"  # x-release-please-version
 description = "LaunchDarkly SDK for AI"
 authors = [{name = "LaunchDarkly", email = "dev@launchdarkly.com"}]
 license = {text = "Apache-2.0"}

packages/sdk/server-ai/src/ldai/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.16.1"  # x-release-please-version
+__version__ = "0.17.0"  # x-release-please-version
 
 from ldclient import log