fix(reasoning-cache): include xiaomi-mimo in replay provider/model detection

[NomenAK]

Summary

Xiaomi MiMo (xiaomi-mimo provider, mimo-v2.5-pro / mimo-v2.5 / mimo-v2-omni models) enforces the same "echo reasoning_content on subsequent turns" contract as DeepSeek and Kimi-thinking. Without replay, the upstream returns 400:

```
data:{"error":{"code":"400","message":"Param Incorrect","param":"The reasoning_content in the thinking mode must be passed back to the API.","type":""}}
```

Repro: client sends a multi-turn /v1/messages body where the assistant history has tool_use blocks but no thinking blocks (Capy and most Anthropic-SDK BYOK clients strip thinking on the wire). MiMo refuses the call because the previous assistant turn's reasoning_content is missing.

The reasoning replay cache (issue #1628) already captures reasoning_content from non-streaming responses with tool_calls (chatCore.ts:4080) and re-injects it on the request side (translator/index.ts:212-242). But the gate requiresReasoningReplay(provider, model) did not include MiMo, so the captured reasoning was discarded on the next turn instead of replayed.

Fix

• Add `"xiaomi-mimo"` to `REASONING_REPLAY_PROVIDERS`
• Add `/^mimo[-.]?v\d/i` to `REASONING_REPLAY_MODEL_PATTERNS` (defensive match if a wildcard route assigns a non-`xiaomi-mimo` provider ID to a `mimo-*` model alias)

Test plan

• 4 new test cases in `reasoning-cache.test.ts` covering both provider-id and model-pattern detection paths, including `XIAOMI-MIMO` uppercase normalization (40/40 green)
• Smoke-validated against live MiMo via OmniRoute multi-turn call — 400 disappears, MiMo accepts the replayed reasoning_content

🤖 Generated with Claude Code

[gemini-code-assist]

Code Review

This pull request adds support for Xiaomi MiMo reasoning models by updating the reasoning cache service to include the xiaomi-mimo provider and a regex pattern for MiMo models. It also includes unit tests to verify the detection logic. A review comment suggests removing the start-of-string anchor from the new regex pattern to maintain consistency with existing patterns and ensure robustness against prefixed model IDs.

[gemini-code-assist]

The regex for mimo models uses a start-of-string anchor (^), which is inconsistent with the other patterns in REASONING_REPLAY_MODEL_PATTERNS (such as deepseek-r1, qwq, or qwen.*think). If a model ID is prefixed (e.g., provider/mimo-v2.5), this pattern will fail to match. Removing the ^ anchor would maintain consistency with the existing patterns and improve robustness against prefixed model names.

Suggested change

	/^mimo[-.]?v\d/i,
	/mimo[-.]?v\d/i,

[diegosouzapw]

Thanks @NomenAK for the Xiaomi MiMo reasoning replay fix. I synced the PR branch with release/v3.8.0, validated tests/unit/reasoning-cache.test.ts locally, and will include this in the upcoming release.