[managed] Enable OpenAI managed-proxy fallback routing in provider bootstrap#26211
Merged
noanflaherty merged 1 commit intoApr 17, 2026
Conversation
![devin-ai-integration[bot]](https://avatars.githubusercontent.com/in/811515?s=60&v=4){kind=small}
Comment on lines
29
to
33
| openai: { | ||
| name: "openai", | ||
| managed: false, | ||
| managed: true, | ||
| proxyPath: "/v1/runtime-proxy/openai", | ||
| }, |
Contributor
There was a problem hiding this comment.
🚩 OpenAI SDK baseURL behavior with proxy path is correct but worth noting
When using the managed proxy, the OpenAI SDK's baseURL is set to something like https://platform.example.com/v1/runtime-proxy/openai. The SDK appends its own API paths (e.g. /responses) to this base, yielding https://platform.example.com/v1/runtime-proxy/openai/responses. This matches how the Anthropic SDK handles its baseURL for the anthropic managed proxy. The platform proxy server must be configured to handle this path structure — but that's a server-side concern in vellum-assistant-platform, not a bug here.
Was this helpful? React with 👍 or 👎 to provide feedback.
noanflaherty
merged commit 9c3f06f
into
noanflaherty/managed-openai-native-web-search
12 checks passed
noanflaherty
added a commit
that referenced
this pull request
Apr 17, 2026
* [managed] Enable OpenAI managed-proxy fallback routing in provider bootstrap (#26211) * [openai] Route inference-provider-native web search to Responses native web_search tool (#26212) * [macOS] Allow managed inference provider selection beyond Anthropic (#26218) * [macOS] Allow managed inference provider selection beyond Anthropic * fix: capture draftProvider before async Task to prevent race condition Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * [macOS] Allow provider-native web search when managed inference provider supports it (#26230) * [macOS] Allow provider-native web search when managed inference provider supports it * fix: gate mode-specific web search invalidation on modeChanging to prevent false-positive alerts Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * fix: generalize alert message to cover both mode and provider changes * fix: scope provider-native invalidation to your-own web search mode --------- Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * fix: emit server_tool_start/complete events for OpenAI native web search (#26240) * fix: render web_search_call output items in LLM context diagnostics (#26243) * fix: add server_tool_use content blocks and tests for OpenAI native web search (#26246) * fix: diagnostics polish — web_search_call tests, tool count fix, loop consistency (#26250) * fix: emit paired web_search_tool_result blocks to prevent repairHistory corruption The OpenAI Responses provider emitted server_tool_use content blocks for web_search_call items but did not emit matching web_search_tool_result blocks. repairHistory() treats any unpaired server_tool_use as an interrupted search and injects a synthetic web_search_tool_result_error, which corrupts conversation history by making successful searches appear as failures. After each server_tool_use block, also emit a paired web_search_tool_result with empty content (since OpenAI weaves search results into the text output). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
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
1 participant
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
Part of plan: managed-openai-native-web-search.md (PR 1 of 4)