fix(cursor): send Grok 4.5 Fast parameters - #881
Closed
jbaehova wants to merge 3 commits into
Closed
Conversation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Plus Run ID: 📒 Files selected for processing (4)
📝 WalkthroughWalkthroughCursor Grok Fast routing now sends ChangesCursor Grok Fast routing
Estimated code review effort: 3 (Moderate) | ~20 minutes Sequence Diagram(s)sequenceDiagram
participant Caller
participant createCursorRequest
participant protobuf-request
participant Cursor
Caller->>createCursorRequest: select cursor/grok-4.5-fast
createCursorRequest->>protobuf-request: pass grok-4.5 and requested parameters
protobuf-request->>Cursor: send requestedModel with effort and fast=true
protobuf-request->>Cursor: omit modelDetails when explicit parameters exist
Possibly related PRs
Suggested reviewers: 🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs-site/src/content/docs/reference/adapters.md`:
- Around line 171-172: Update the translated Cursor adapter sections in the ja,
ko, ru, and zh-cn adapter reference documents to state that cursor/grok-4.5-fast
remains selectable while the request sends canonical grok-4.5 with separate
effort and fast=true parameters.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro Plus
Run ID: 672d6a48-7c92-4494-8597-435bd8f459a8
📒 Files selected for processing (8)
docs-site/src/content/docs/reference/adapters.mdsrc/adapters/cursor/effort-map.tssrc/adapters/cursor/protobuf-request.tssrc/adapters/cursor/request-builder.tssrc/adapters/cursor/types.tsstructure/04_transports-and-sidecars.mdtests/cursor-blob.test.tstests/cursor-effort-suffix.test.ts
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs-site/src/content/docs/ja/reference/adapters.md`:
- Around line 107-108: Align the translated Cursor adapter documentation with
the wire contract by identifying requested_model.parameters as the container for
effort and fast=true in docs-site/src/content/docs/ja/reference/adapters.md
lines 107-108, docs-site/src/content/docs/ko/reference/adapters.md lines
126-127, docs-site/src/content/docs/ru/reference/adapters.md lines 137-138, and
docs-site/src/content/docs/zh-cn/reference/adapters.md lines 120-121; preserve
the existing localized wording while adding the localized equivalent of the
field name at each site.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro Plus
Run ID: c3c7f1c9-362a-4741-af6f-1574e4affd5c
📒 Files selected for processing (4)
docs-site/src/content/docs/ja/reference/adapters.mddocs-site/src/content/docs/ko/reference/adapters.mddocs-site/src/content/docs/ru/reference/adapters.mddocs-site/src/content/docs/zh-cn/reference/adapters.md
Owner
|
Landed in #892 as 0793edc + 05bb74f + ab6a1ed (authorship preserved). Adversarial review of the stacked branch found no issues in the requested_model path — the model_details/requested_model exclusivity and the discovery-gating were verified against the wt2 cursor blob/frame work. Closing as superseded; thanks for the corrective follow-up to #797. |
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
2 participants
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
cursor/grok-4.5-fastas the Codex-facing selector, but send Cursor's canonicalgrok-4.5model with separateeffortandfast=trueparametersAgentRunRequest.requested_modelwithout also sending the conflicting legacymodel_detailsWhy
This is a corrective follow-up to my earlier PR #797, which added
cursor/grok-4.5-fastsupport and was included in OpenCodex 2.9.1. After the release, I tested the model through OpenCodex and found that the Fast selector was visible but every request returned 502, whilecursor/grok-4.5continued to work. The earlier PR correctly fixed discovery, but it also assumed Cursor accepted the flattenedgrok-4.5-{effort}-fastid as the outgoing model selection. This PR corrects that request-side assumption.Current Cursor clients represent this selection as base model
grok-4.5plus model parameters (effort=<tier>,fast=true). Cursor's generated protobuf already exposes the same string-valuedRequestedModel.parameterschannel, and the current official SDK leaves legacymodel_detailsunset when it sends a requested model. Using that contract also resolves the ordering concern raised in the automated review on #797: discovered flattened ids now gate availability only and no longer determine the outgoing request shape.The working non-Fast Grok path remains unchanged and continues to use its existing flat effort id. Cursor Router behavior is also unchanged: its already-live
model_details + requested_model(optimization)shape remains intact.Protocol evidence
@cursor/sdk1.0.24 models selections as{ id, params }; its local runtime serializes that shape intoRequestedModel.model_idand string-valuedparameters.{ id: "fast", value: "false" }instead of treating the resolved Fast slug as the request model id.effortandfast; the live probes below verify the exact Grok 4.5 combination used by this patch.User impact
Selecting
cursor/grok-4.5-fastnow reaches Cursor Grok 4.5 Fast instead of failing before generation. Low, medium, high, clamped xhigh, and omitted-effort selections all preserve their existing picker behavior; only the Cursor wire representation changes.Verification
bun test tests/cursor-discovery.test.ts tests/cursor-effort-suffix.test.ts tests/cursor-blob.test.ts tests/cursor-request-builder.test.ts tests/cursor-adapter.test.ts— 111 passedbun test tests/cursor-effort-suffix.test.ts tests/cursor-blob.test.ts— 61 passed after the final regression assertionbun run prepushcd docs-site && bun run buildgit diff --checkcursor-agent 2026.07.23-e383d2blive probe:--model 'grok-4.5[effort=high,fast=true]'completed withOKgrok-4.5witheffort=high,fast=trueand completed withdone / OKChecklist
Summary by CodeRabbit
New Features
cursor/grok-4.5-fast.Bug Fixes
Documentation