feat(TinyGraphRAG): add MiniMax as LLM provider

[octo-patch]

Summary

Add MiniMax as a third LLM provider for TinyGraphRAG, alongside ZhipuAI and Groq.

MiniMax provides an OpenAI-compatible Chat Completions API, so the implementation reuses the openai SDK with a custom base_url—no additional dependencies are needed.

Supported models

Model	Description
MiniMax-M2.7	Default model, 204K context window
MiniMax-M2.7-highspeed	Same performance, faster and more agile

Changes

• tinygraph/llm/minimax.py — minimaxLLM class extending BaseLLM, follows the same pattern as groqLLM and zhipuLLM
• tests/test_minimax_llm.py — 13 unit tests (init, predict, error handling)
• tests/test_minimax_integration.py — 3 integration tests against the live MiniMax API
• readme.md — Added MiniMax usage example in the LLM module section
• help.ipynb — Added commented MiniMax configuration example

Usage

from tinygraph.llm.minimax import minimaxLLM

llm = minimaxLLM(
    model_name="MiniMax-M2.7",
    api_key="your-minimax-api-key",
)
print(llm.predict("Hello, how are you?"))

Test plan

• 13 unit tests pass (mocked OpenAI client)
• 3 integration tests pass (live MiniMax API with MINIMAX_API_KEY)
• Existing code unchanged — fully additive changes only