test(cursor): verify Grok 4.6 routing

[iamsupersocks]

Summary

• Replace the pre-release Grok 4.6 comments with launch-day provenance from the official xAI announcement and an account-verified cursor-grok-4.6-high run.
• Add regression coverage for Cursor live-discovery matching, 500k context metadata, the account-observed cursor-grok-4.6-high wire ID, the announced low/medium/high ladder, and the inherited parameterized Fast encoding.
• Keep the existing Grok 4.6 runtime behavior unchanged.

Source: https://x.ai/news/grok-4-6

Verification

• bun install --frozen-lockfile
• bun test tests/cursor-discovery.test.ts tests/cursor-effort-suffix.test.ts — 23 passed, 0 failed
• bun run typecheck
• bun run test — 11,388 passed, 11 skipped, 0 failed
• GitHub target, hygiene, and label checks are green

Checklist

• Scope stays focused and avoids unrelated cleanup.
• Docs or release notes were updated when needed. (Not needed: this PR verifies already-integrated runtime behavior.)
• Security-sensitive changes were reviewed for secrets, auth, and unsafe defaults. (Tests and comments only; no credentials or auth behavior changed.)

Review readiness checklist

This PR stays in draft until every box below is ticked. Tick all four boxes once the requirements are met:

• All CI tests are green on my local testing.

• I pushed my PR to the latest dev commit.

• I resolved all correct Codex and CodeRabbit findings.

• My PR is ready for review.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 8e9ba3b3-25e5-4166-9749-8c00aec5715f

📥 Commits

Reviewing files that changed from the base of the PR and between c125b61 and 13c7b1f.

📒 Files selected for processing (4)

• src/adapters/cursor/discovery.ts
• src/adapters/cursor/effort-map.ts
• tests/cursor-discovery.test.ts
• tests/cursor-effort-suffix.test.ts

---

📝 Walkthrough

Walkthrough

The change documents verified Cursor Grok 4.6 support and expands tests for model discovery, context windows, reasoning efforts, wire IDs, Fast variants, and effort clamping.

Changes

Cursor Grok 4.6 support

Layer / File(s)	Summary
Model discovery coverage src/adapters/cursor/discovery.ts, tests/cursor-discovery.test.ts	Updates the Grok 4.6 catalog note. Tests cover static inclusion, live wire-ID formats, filtering, context windows, and low/medium/high reasoning efforts.
Effort suffix handling src/adapters/cursor/effort-map.ts, tests/cursor-effort-suffix.test.ts	Updates the Grok 4.6 effort-tier note. Tests cover regular and Fast request IDs, effort clamping, Fast parameters, and supported effort ladders.

Estimated code review effort: 2 (Simple) | ~10 minutes

Possibly related PRs

• lidge-jun/opencodex#646: Modifies Cursor model discovery and effort-suffixed model ID tests.
• lidge-jun/opencodex#730: Extends Cursor Grok discovery and effort-suffix handling.
• lidge-jun/opencodex#797: Updates Cursor Grok effort mapping, wire IDs, discovery logic, and tests.

Suggested reviewers: lidge-jun

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly identifies the Cursor test changes that verify Grok 4.6 routing, which matches the main purpose of the pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

✅ Deterministic PR hygiene checks passed.

[github-actions]

⏳ DRAFT

• review readiness checklist open (2/4 boxes ticked).

What to do

• Tick all four boxes in the PR description once you're done (currently 2/4).

Review readiness checklist

• ✅ All CI tests are green on my local testing.
• ⬜ I pushed my PR to the latest dev commit.
• ✅ I resolved all correct Codex and CodeRabbit findings.
• ⬜ My PR is ready for review.

2/4 boxes ticked.

This PR stays in draft until every box above is ticked.

[iamsupersocks]

@coderabbitai review

[coderabbitai]

⚠️ Action not completed

Already reviewed.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.

[coderabbitai]

Tip

For best results, initiate chat on the files or code changes.

@github-actions[bot] The readiness gate reports that all four checklist items are complete for head SHA 13c7b1f89f3e5bbd1f29d8c92efd407c3715ce45. The pull request is marked ready for review.

---

🧠 Learnings used

Learnt from: Wibias
Repo: lidge-jun/opencodex PR: 986
File: .github/workflows/enforce-pr-target.yml:631-636
Timestamp: 2026-08-04T06:41:30.089Z
Learning: For the PR readiness gate in `.github/workflows/enforce-pr-target.yml`, the required scope is to bind contributor checklist completion and maintainer notification to the exact PR head SHA, then reset them after a new head commit. Maintainer approval staleness is separate from this gate and should be enforced by GitHub branch protection, such as `dismiss-stale-reviews` or `require_last_push_approval`, when configured for `dev`.

_{You are interacting with an AI system.}

[coderabbitai]

Tip

For best results, initiate chat on the files or code changes.

@github-actions[bot] The readiness gate is complete for head SHA 13c7b1f89f3e5bbd1f29d8c92efd407c3715ce45. The pull request is marked ready for review.

---

🧠 Learnings used

Learnt from: Wibias
Repo: lidge-jun/opencodex PR: 986
File: .github/workflows/enforce-pr-target.yml:631-636
Timestamp: 2026-08-04T06:41:30.089Z
Learning: For the PR readiness gate in `.github/workflows/enforce-pr-target.yml`, the required scope is to bind contributor checklist completion and maintainer notification to the exact PR head SHA, then reset them after a new head commit. Maintainer approval staleness is separate from this gate and should be enforced by GitHub branch protection, such as `dismiss-stale-reviews` or `require_last_push_approval`, when configured for `dev`.

_{You are interacting with an AI system.}

[Wibias]

The runtime behavior is unchanged and CI is green, but the new provenance test currently overstates what was actually verified. Please separate observed Cursor wire IDs from inferred/announced tier behavior so the regression suite does not manufacture evidence for low/medium (or Fast transport) that the PR says was not directly account-verified.

The provenance blocker was addressed in c53d218: the recorded Grok 4.6 fixture now contains only the account-observed high wire ID, with low/medium and Fast explicitly separated as announced/inherited behavior rather than recorded wire evidence. Clearing the old changes-requested state; final approval still waits for rebase onto current dev and exact-head CI.

[Wibias]

Re-review on exact head c53d218: the previous provenance blocker is resolved. The recorded Grok 4.6 fixture now contains only the account-observed cursor-grok-4.6-high; low/medium and Fast are clearly separated as announced/inherited behavior rather than claimed wire captures. I found no new correctness issue in the four-file diff. I’m not approving this head yet because the PR is still draft and based on c125b61 while current dev is 75cacdf (38 commits ahead). None of the four PR files changed across that dev advance, so the rebase should be mechanical, but please rebase and let exact-head CI run before final approval.