Repository navigation
Integrate dynamic profiles with OpenAI Tool loops - #344
Merged
Merged
Conversation
…dynamic-profiles # Conflicts: # Sources/AnyLanguageModel/LanguageModelSession.swift # Sources/AnyLanguageModel/SessionProperties.swift # Sources/AnyLanguageModelMacros/SessionPropertyEntryMacro.swift # Tests/AnyLanguageModelTests/APICompatibilityAnyLanguageModelTests.swift # Tests/AnyLanguageModelTests/APICompatibilityFoundationModelsTests.swift # Tests/AnyLanguageModelTests/SessionPropertyTests.swift
…dynamic-profiles # Conflicts: # Sources/AnyLanguageModel/SessionProperties.swift # Tests/AnyLanguageModelTests/SessionPropertyTests.swift
…dynamic-profiles # Conflicts: # Sources/AnyLanguageModel/LanguageModelSession.swift
…ofile-tool-continuations # Conflicts: # Sources/AnyLanguageModel/DynamicProfile.swift # Sources/AnyLanguageModel/LanguageModelSession.swift
…ile-tool-continuations
…stream-session-properties
…stream-dynamic-profiles
…stream-profile-tool-continuations
Take main for everything that PR 315 and PR 377 own and keep only the OpenAI tool loop integration from this branch. Send PR 377's schema text through the profile's tool continuation messages. Keep the new resolution's callback history and history version when a request context shares callback state with the prepared request. Without them, each callback history write after the first tool round reported a conflict. Count the profile's instructions entry, which main records, in the handoff fixtures.
mattt
added a commit
to qoli/AnyLanguageModel
that referenced
this pull request
Nov 5, 2026
main has the shared profile runtime from huggingface#315 and the work in huggingface#344 to huggingface#348, so main's version is kept for all files except SystemLanguageModel.swift and NativeProfileHistoryTests.swift. The tests now expect what main does when it creates a profile session: it evaluates the profile once and records an instructions entry.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Scope
OpenAI Chat Completions / Responses integration for #315's shared Compatibility SPI. Provider PRs are siblings, not a chain.
Depends on #315 cf4ad0d, transitively #314/#313 and main's #352. Isolated layer comparison.
Both APIs retain upstream #325 stable/plain JSON replay and assistant-content metadata. Each complete ToolCalls entry is created before hooks/Tool execution and committed unchanged.
Upstream 1.1 synchronization
Normal history-preserving merges synchronize upstream main 7d43219, including #357 per-request usage accounting, #341 Ollama reasoning, #358 host-scoped Chat Completions reasoning, #300 response-start timeouts, #362 cancellation collection, #364 partial structured snapshots and #366/#367 bounds/reasoning/system-Tool fixes. Earlier #336 final-request usage, #340 HTTPTransport, #297/#325/#331/#343/#352/#354/#356 remain intact. No force-push, new checkout, replacement PR or extra fallback.
The continuation mapper carries the producing API host into reasoning replay. A Tool round's reasoning is retained once before its stable ToolCalls entry and dispatched to that producing request's profile before Tool callbacks can switch profiles. Ordinary/profile and streaming/non-streaming request-history regressions first failed, then passed; a separate before/after-profile-switch regression reproduced both callback failures before the dispatch fix. Upstream host filtering, stable/plain JSON arguments, final-request usage, injected transport and timeout semantics are preserved.
Fresh local verification (2026-10-23)
Head 5e95c10.
CI=1 GIT_CONFIG_GLOBAL=/dev/null swift test: 957 tests / 100 suites passed. The separate upstream macro test target also passed 2 tests / 1 suite. Strict formatting/diff checks passed. These are compile/offline tests, not live API, model-weight or Apple Intelligence inference acceptance.The offline suites retain streaming/non-streaming history, callback, reused-ID and handoff accounting regressions. Constructor visibility is still SPI; no new PR or force-push.