chore: release main

[github-actions]

🤖 I have created a release beep boop

launchdarkly-server-sdk-ai: 0.17.0

0.17.0 (2026-03-26)

⚠ BREAKING CHANGES

• Split track_metrics_of into sync and async variants (#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)

Features

• Add get_ai_usage_from_response to openai_helper (4fab18f)
• Add get_tool_calls_from_response and sum_token_usage_from_messages to langchain_helper (4fab18f)
• Add ModelRunner ABC with invoke_model() and (453c71c)
• add optimization package stub (872e81e)
• Add optional graph_key to all LDAIConfigTracker track_* methods for graph correlation (4fab18f)
• Add track_tool_call/track_tool_calls to LDAIConfigTracker (4fab18f)
• Adds optimization package stub (58b7731)
• Deprecated Chat object in favor of ManagedModel (453c71c)
• Deprecated create_chat(), use create_model() on the LDAIClient (453c71c)
• Extract shared utilities to langchain_helper (453c71c)
• Extract shared utilities to openai_helper (453c71c)
• Introduce ManagedModel and ModelRunner (PR-3) (#104) (453c71c)
• Restructure provider factory and support additional create methods (#102) (e6e4907)
• Split track_metrics_of into sync and async variants (#112) (4fab18f)

Bug Fixes

• make AIGraphTracker.track_total_tokens accept Optional[TokenUsage], skip when None or total <= 0 (4fab18f)
• Remove node-scoped methods from AIGraphTracker (track_node_invocation, track_tool_call, track_node_judge_response), use related AIConfigTracker methods instead (4fab18f)
• use time.perf_counter_ns() for sub-millisecond precision in duration calculations (4fab18f)

launchdarkly-server-sdk-ai-langchain: 0.4.0

0.4.0 (2026-03-26)

⚠ BREAKING CHANGES

• Split track_metrics_of into sync and async variants (#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)

Features

• Add get_ai_usage_from_response to openai_helper (4fab18f)
• Add get_tool_calls_from_response and sum_token_usage_from_messages to langchain_helper (4fab18f)
• Add ModelRunner ABC with invoke_model() and (453c71c)
• add optimization package stub (872e81e)
• Add optional graph_key to all LDAIConfigTracker track_* methods for graph correlation (4fab18f)
• Add track_tool_call/track_tool_calls to LDAIConfigTracker (4fab18f)
• Adds optimization package stub (58b7731)
• Deprecated Chat object in favor of ManagedModel (453c71c)
• Deprecated create_chat(), use create_model() on the LDAIClient (453c71c)
• Extract shared utilities to langchain_helper (453c71c)
• Extract shared utilities to openai_helper (453c71c)
• Introduce ManagedModel and ModelRunner (PR-3) (#104) (453c71c)
• Restructure provider factory and support additional create methods (#102) (e6e4907)
• Split track_metrics_of into sync and async variants (#112) (4fab18f)

Bug Fixes

• make AIGraphTracker.track_total_tokens accept Optional[TokenUsage], skip when None or total <= 0 (4fab18f)
• Remove node-scoped methods from AIGraphTracker (track_node_invocation, track_tool_call, track_node_judge_response), use related AIConfigTracker methods instead (4fab18f)
• use time.perf_counter_ns() for sub-millisecond precision in duration calculations (4fab18f)

launchdarkly-server-sdk-ai-openai: 0.3.0

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)

Features

• Add ModelRunner ABC with invoke_model() and (453c71c)
• add optimization package stub (872e81e)
• Adds optimization package stub (58b7731)
• Deprecated Chat object in favor of ManagedModel (453c71c)
• Deprecated create_chat(), use create_model() on the LDAIClient (453c71c)
• Extract shared utilities to langchain_helper (453c71c)
• Extract shared utilities to openai_helper (453c71c)
• Introduce ManagedModel and ModelRunner (PR-3) (#104) (453c71c)
• Restructure provider factory and support additional create methods (#102) (e6e4907)

launchdarkly-server-sdk-ai-optimization: 0.1.0

0.1.0 (2026-03-26)

Features

• Add optimization package stub (#109) (ebd5166)
• Adds optimization package stub (58b7731)
• Adds optimization package stub (cc85a05)

Bug Fixes

• consistency with other makefiles (b9a5601)

---

This PR was generated with Release Please. See documentation.

---

Note

Medium Risk
Primarily a Release Please version/changelog update, but it ships breaking API changes (tracking method splits/removals and provider factory restructuring) that can break downstream consumers if they upgrade.

Overview
Publishes a new release across packages: launchdarkly-server-sdk-ai -> 0.17.0, LangChain provider -> 0.4.0, OpenAI provider -> 0.3.0, and introduces the optimization package at 0.1.0 (manifest + pyproject.toml updates).

Updates generated changelogs/provenance/version constants to reflect the release, with notes for breaking changes (e.g., track_metrics_of split into sync/async variants, removal of node-scoped AIGraphTracker methods in favor of AIConfigTracker, and provider factory restructuring) and related feature/bugfix callouts.

^{Written by Cursor Bugbot for commit 846fe7d. This will update automatically on new commits. Configure here.}

[jsonbailey]

Hold off on merging this. I am working to merge in all of the Agent Graph work so that we don't have multiple breaking change versions.