Add MiniMax as a first-class LLM provider

[octo-patch]

Summary

Add MiniMax AI as a new LLM provider alongside OpenAI, Anthropic, and Vertex AI. MiniMax offers an OpenAI-compatible API with models including MiniMax-M2.7 (latest, 1M context), MiniMax-M2.5, and MiniMax-M2.5-highspeed (204K context).

Changes

• LLM layer (skllm/llm/minimax/): credentials, completion (with temperature clamping to [0,1] and think-tag stripping), and mixin modules following the existing Anthropic/GPT pattern
• Model layer (skllm/models/minimax/): zero-shot/few-shot/CoT classifiers, multi-label classifiers, dynamic few-shot classifiers, summarizer, translator, and NER
• Config (skllm/config.py): SKLLMConfig.set_minimax_key() / get_minimax_key() for API key management
• Model constants (skllm/model_constants.py): MINIMAX_MODEL default
• README (README.md): Quick start example for MiniMax usage

Usage

Test plan

• 27 unit tests covering mixins, completion, credentials, config, model constants, and all model class constructors
• 6 integration tests verifying live API calls (basic completion, system message, JSON response, text completion mixin, classifier mixin, config)
• Existing GPT, Anthropic, and utility tests pass without regressions

Files changed: 15 files, 1089 additions