fix(cursor): expose Grok 4.6 xhigh thinking

[Melvynx]

Summary

• Extend the existing Cursor Grok 4.6 effort ladder with the xhigh tier advertised by the Cursor Agent CLI.
• Map cursor/grok-4.6[effort=xhigh] to Cursor wire model cursor-grok-4.6-xhigh.
• Add a regression test for the wire ID and the Codex-facing picker ladder.

This intentionally complements, rather than duplicates, #1547, which verifies the existing low/medium/high Grok 4.6 support.

Verification

• bun test tests/cursor-effort-suffix.test.ts tests/cursor-discovery.test.ts — 21 passed
• bun run typecheck — passed
• Live local OpenCodex probe through authenticated Cursor: cursor/grok-4.6 returned HTTP 200 for low, high, and xhigh.

Checklist

• Scope stays focused and avoids unrelated cleanup.
• Docs or release notes were updated when needed. No public docs change: this only adds an advertised picker tier.
• Security-sensitive changes were reviewed for secrets, auth, and unsafe defaults. No auth or transport 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.

[github-actions]

✅ Deterministic PR hygiene checks passed.

[coderabbitai]

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 896d0418-870c-48a3-9845-0db2ee77b86f

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

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]

⏳ DRAFT

• review readiness checklist open (0/4 boxes ticked).

What to do

• Tick all four boxes in the PR description once you're done (currently 0/4).

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.

0/4 boxes ticked.

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

[Ingwannu]

The regular Grok 4.6 direction is supported on exact head f2ee5d1973192b7daa20942c7af333baf01004b0: the two focused suites pass 21/21, typecheck/privacy/diff check are green, and the PR reports authenticated HTTP 200 probes for regular cursor/grok-4.6 at low/high/xhigh.

I am requesting changes because the code also adds xhigh to grok-4.6-fast, but the evidence and new regression cover only the regular model. That materially changes the parameterized Fast request from clamping xhigh -> high to sending { effort: \"xhigh\", fast: \"true\" }. PR #1547 deliberately separated observed regular wire IDs from inherited Fast behavior; this PR should preserve that evidence boundary.

Please do one of the following:

1. limit the tier change to grok-4.6 and leave grok-4.6-fast at low/medium/high; or
2. provide an authenticated Fast xhigh success capture and add a regression for selectionFor(\"cursor/grok-4.6-fast\", \"xhigh\") plus its advertised picker ladder.

After that, rebase onto the latest dev, complete the readiness checklist, and obtain exact-head CI. The regular-model change remains small and valuable.

[Melvynx]

Closing: this only duplicated the already-open Grok 4.6 integration work in #1547. The local setup will track that PR branch instead.