preserve OpenAI parallel tool-call policy - #97
Closed
raine wants to merge 1 commit into
Closed
Conversation
OpenAI-compatible requests accepted serial tool execution without carrying that constraint into provider translation, and rejected the standard parallel value. This could silently reverse caller intent on non-Codex provider routes. Encode the OpenAI setting in the established Anthropic tool-choice policy and share its interpretation across Codex, Kimi, and Grok translators. Kimi and Grok propagate the resolved value to their upstream APIs, while native Codex Responses retains its passthrough behavior. Cover both OpenAI routes, tool-choice modes, provider wire requests, validation, and native Codex passthrough. Document accepted values and precedence.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
parallel_tool_callson shared OpenAI-compatible ingressDesign
OpenAI
parallel_tool_callsis the inverse of Anthropictool_choice.disable_parallel_tool_use. The shared parser attaches that policy to the translated tool choice, synthesizingautowhen the caller omits a choice. An explicit OpenAI value determines the translated policy without changingauto,none,required, or named-function selection.Provider translators read the shared policy. Kimi and Grok emit the corresponding optional upstream
parallel_tool_callsfield. Codex uses the same reader for Anthropic Messages translation, while native/v1/responsescontinues forwarding the original request body.Verification
The pre-commit Checkle hook passed:
Regression coverage includes both OpenAI routes, omitted and explicit tool choices, both boolean values, invalid values, Kimi and Grok translated upstream payloads, and native Codex passthrough.