feat: add MiniMax AI as alternative LLM provider

[octo-patch]

Summary

Add MiniMax AI (M2.7 and M2.7-highspeed, 204K context) as an alternative LLM provider alongside Google Gemini. This is the first step toward making All Model Chat truly multi-provider.

Key changes:

• OpenAI-compatible API client (services/api/openaiCompatApi.ts): Lightweight stream + non-stream chat via native fetch() - no new npm dependencies
• Provider routing in useApiInteraction.ts: Detects MiniMax models and routes through the OpenAI-compat path
• MiniMax models added to default model picker (pinned): MiniMax-M2.7, MiniMax-M2.7-highspeed
• Settings UI: Separate MiniMax API key input in Settings > API Configuration
• Temperature clamping: MiniMax requires (0.0, 1.0] - auto-clamped
• Thinking tag stripping: M2.7 <thinking> blocks extracted and displayed properly
• Bilingual i18n: English + Chinese translations for all new UI strings
• 32 unit + integration tests with vitest (provider constants + API client)

Files changed (15 files, ~1116 additions):

File	Change
constants/providerConstants.ts	NEW - MiniMax models, detection helpers
services/api/openaiCompatApi.ts	NEW - OpenAI-compat chat API
tests/providerConstants.test.ts	NEW - 12 unit tests
tests/openaiCompatApi.test.ts	NEW - 20 unit + integration tests
types/settings.ts	Add minimaxApiKey to AppSettings
constants/appConstants.ts	Default minimaxApiKey: null
utils/modelHelpers.ts	Include MiniMax models in defaults
hooks/.../useApiInteraction.ts	MiniMax routing logic
hooks/.../useStandardChat.ts	Bypass Gemini key check for MiniMax
components/.../ApiConfigSection.tsx	MiniMax API key UI
components/.../SettingsContent.tsx	Pass MiniMax props
utils/translations/settings/api.ts	i18n strings
vite.config.ts	Add vitest config
package.json	Test scripts
README.md	Multi-provider docs

How to use

1. Enter your MiniMax API key in Settings > API Configuration > MiniMax AI Configuration
2. Select MiniMax M2.7 or MiniMax M2.7 Highspeed from the model picker
3. Chat as usual - streaming and non-streaming both work

Get your API key at platform.minimax.chat

Test plan

• All 32 tests pass (npm test)
• Build succeeds (npm run build)
• Manual test: Configure MiniMax API key in settings
• Manual test: Select MiniMax model and send a message (streaming)
• Manual test: Verify temperature clamping and thinking tag display
• Manual test: Verify existing Gemini functionality is unaffected