register gpt-6 model tiers in the picker

[damienrj]

Category: improvement
User Impact: GPT-6 Astra, Sol, and Luna appear with clear names and sensible ordering when a connected provider offers them.
Problem: Newly discovered GPT-6 models used generic labels and could appear below older models in the picker.
Solution: Recognize the three official GPT-6 IDs in raw provider results, format Goose model names consistently, and prioritize the newest GPT tier without changing provider discovery or defaults.

File changes

src/features/providers/lib/humanizeModelId.ts
Show official names for raw GPT-6 Astra, Sol, and Luna IDs.

src/features/providers/lib/humanizeModelId.test.ts
Verify the three raw IDs use their intended names.

src/features/providers/lib/modelRecommendations.ts
Format Goose GPT model names and order GPT-6 tiers ahead of older GPT versions while preserving per-family recommendations.

src/features/providers/lib/modelRecommendations.test.ts
Verify GPT-6 naming, ordering, recommendations, and the featured model.

Validation: just check; 20 focused Vitest tests; commit and push hooks, including Tauri check and Clippy.

[morgmart]

🤖 Automated code review

REQUEST_CHANGES. The GPT-6 update broadly promotes any future gpt-* family at version 6 or later, allowing an unrecognized discovered model to outrank the three intended tiers and become the featured or automatically selected model. Supplied GitHub evidence was inspected; several required checks were still in progress at capture time and continue to govern merge readiness.

Deterministic publication result: 1 blocking and 0 non-blocking inline finding(s) publishable; 0 duplicate(s) suppressed; 0 blocking screenshot-evidence requirement(s) in this review body.

[damienrj]

🤖 The requested GPT-6 ranking fix is in b80763f, with an unknown-tier regression test. I replied in the thread and resolved it; all checks now pass. Please re-review the latest commit when available.

[morgmart]

🤖 Automated code review

APPROVE. The previous blocking issue is fixed: only the recognized GPT-6 tiers receive top priority, their order is explicit, and an unknown higher-version GPT family remains mid-ranked and unfeatured. The full current comparison has no new publishable findings. Supplied GitHub evidence is structurally valid and shows all captured checks passing.

Deterministic publication result: 0 blocking and 0 non-blocking inline finding(s) publishable; 0 duplicate(s) suppressed; 0 blocking screenshot-evidence requirement(s) in this review body.

Pending checks: 1 check(s) are not complete.

This approval reflects the completed code review only; merge readiness remains governed by the repository's required checks.