feat: add MiniMax as alternative LLM provider

[octo-patch]

Summary

Add MiniMax (M2.7 / M2.7-highspeed, 204K context) as an alternative LLM provider alongside the existing Qwen (DashScope) integration.

• Uses MiniMax OpenAI-compatible API via langchain4j-open-ai, so no custom SDK needed
• Provider selection via ai.provider property in application.yml (qwen default, minimax available)
• @ConditionalOnProperty + @primary beans ensure clean provider switching without conflicts
• Existing Qwen behavior is 100% unchanged when ai.provider=qwen (default)

Files changed (9 files, 352 additions)

File	Change
pom.xml	Add langchain4j-open-ai 1.1.0 dependency
MiniMaxChatModelConfig.java	New config: ChatModel + StreamingChatModel via OpenAI-compat API
QwenChatModelConfig.java	Add @ConditionalOnProperty + @primary
AiCodeHelperServiceFactory.java	Switch to type-based @Autowired injection
AiCodeHelper.java	Switch to type-based @Autowired injection
application.yml	Add ai.provider property + MiniMax config section
README.md	Add MiniMax in docs, architecture diagram, and credits
MiniMaxChatModelConfigTest.java	9 unit tests
MiniMaxIntegrationTest.java	3 integration tests (require MINIMAX_API_KEY)

How to use

# application.yml
ai:
  provider: minimax
minimax:
  api-key: your-minimax-api-key
  model-name: MiniMax-M2.7        # or MiniMax-M2.7-highspeed
  base-url: https://api.minimax.io/v1

Test plan

• 9 unit tests pass (MiniMaxChatModelConfigTest)
• 3 integration tests pass with live MiniMax API (MiniMaxIntegrationTest)
• Project compiles successfully with Java 21 + Maven
• Verify default ai.provider=qwen still works unchanged
• Verify ai.provider=minimax routes to MiniMax API