feat: add MiniMax as alternative LLM provider

[octo-patch]

Summary

• Add MiniMax as a first-class LLM provider alongside OpenAI
• Support MiniMax-M2.7 and MiniMax-M2.7-highspeed models via OpenAI-compatible API
• Auto-detect provider from available API keys (MINIMAX_API_KEY / OPENAI_API_KEY)
• Add --llm-provider CLI option for explicit provider selection
• Add MiniMaxChatOpenAI subclass that adapts deprecated function-calling kwargs to tools/tool_choice format and strips <think> tags from responses

Changes

File	Description
integuru/util/LLM.py	Provider presets, MiniMaxChatOpenAI subclass, _detect_provider(), updated LLMSingleton
integuru/__main__.py	Add --llm-provider CLI option
integuru/main.py	Pass provider to LLMSingleton.set_provider()
README.md	Provider table, MiniMax setup instructions
tests/test_llm_provider.py	18 unit tests (presets, detection, provider switching, backward compat)
tests/test_minimax_integration.py	3 integration tests (basic invoke, function calling, alternate model)

Usage

Test plan

• All 18 unit tests pass
• All 3 MiniMax integration tests pass (basic invoke, function calling, alternate model)
• Backward compatibility: existing OpenAI behavior unchanged when no provider specified
• Pre-existing test failures (missing test.har fixture) are unrelated to this PR

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}