docs(passthrough): managed ids cover streamed responses and use the real passthrough prefix - #1022
Merged
Conversation
…eal passthrough prefix
Contributor
Author
|
bugbot run |
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Merged
7 tasks
There was a problem hiding this comment.
✅ Bugbot reviewed your changes and found no new issues!
Comment @cursor review or bugbot run to trigger another review on this PR
Reviewed by Cursor Bugbot for commit 39b00b7. Configure here.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Companion to BerriAI/litellm#38320 (passthrough managed ids for streamed responses)
The page previously used
/openai/v1/filesand/openai/v1/batchesas its example routes. On a live proxy those paths are served by the unified native endpoints, which are matched before the generic passthrough router, sopassthrough_managed_object_idsnever sees them and a user following the docs got no protection and no error. Every example now uses the real passthrough prefixes (/openai_passthrough/...,/azure/openai/...) and an admonition spells out that the three/openai/v1/*object routes are native endpoints with their own controlsAlso documents what the linked PR adds and what stays out of scope: streamed
POST /v1/responsesis now owned and rewritten from the firstresponse.createdevent; objects the proxy never saw (created before the flag, or provider-direct) stay unowned, with one custom passthrough endpoint per team carrying that team's own provider key as the strict isolation option; raw provider ids that map to another caller's resource are refused with 404Note
Low Risk
Documentation-only changes with no runtime or security behavior modified in this PR.
Overview
Updates Passthrough Managed IDs documentation so examples match real proxy routing and reflect behavior from the companion implementation PR.
Route corrections: All OpenAI passthrough examples now use
/openai_passthrough/...instead of/openai/v1/..., with an info callout explaining that/openai/v1/files,/openai/v1/batches, and/openai/v1/responsesare native routes (userequire_managed_filesor Responses ownership there, not passthrough managed IDs).Streaming: The comparison table and limitations now state that
POST /v1/responseswithstream: truerewritesresponse.idfrom the firstresponse.createdevent through the stream; other SSE passthrough streams are still relayed unchanged.Ownership edge cases: New section on objects the proxy never recorded (pre-flag or provider-direct), plus per-team custom passthrough endpoints with separate provider keys as the strict isolation pattern. Raw provider IDs are documented as owner-checked when mapped to a managed resource (404 for another tenant’s ID); unrecorded raw IDs still pass through.
Reviewed by Cursor Bugbot for commit 39b00b7. Bugbot is set up for automated code reviews on this repo. Configure here.