feat(models): improve Kilo Auto model descriptions

[kilo-code-bot]

Summary

• Updated descriptions for all four Kilo Auto models to clearly state which underlying models are used and, where applicable, what reasoning mode is applied per mode.
  • Frontier: Now mentions Claude Opus 4.6 (extended thinking) for planning/architecture/debug/general, and Claude Sonnet 4.6 (extended thinking) for coding/exploration.
  • Balanced: Now mentions Kimi K2.5 (with reasoning) for planning/architecture/debug/general, and Minimax M2.5 for coding/exploration.
  • Free: Now explicitly names Minimax M2.5.
  • Small: Now explicitly names GPT-5 Nano and characterises the use case.

Verification

• No logic was changed; only string literals in description fields were updated.
• Ran pnpm typecheck — no errors.

Visual Changes

N/A

Reviewer Notes

These descriptions surface in the model picker UI shown to users. The goal is to give users enough detail to understand routing behaviour without having to read the source code.

[kilo-code-bot]

Code Review Summary

Status: No Issues Found | Recommendation: Merge

Files Reviewed (1 files)

• src/lib/kilo-auto-model.ts

---

_{Reviewed by gpt-5.4-20260305 · 154,085 tokens}