feat: add MiniMax as first-class LLM provider

[octo-patch]

Summary

Add MiniMax as a first-class LLM provider in ValueCell, supporting MiniMax-M2.7, M2.7-highspeed, M2.5, and M2.5-highspeed models (all with 204K context length) via the OpenAI-compatible API.

Changes

Backend (6 files)

• MiniMaxProvider class in factory.py — extends ModelProvider, uses agno's OpenAILike with temperature clamping to (0.0, 1.0] range
• Registered in ModelFactory._providers registry
• Provider YAML config (minimax.yaml) with 4 models and connection settings
• config.yaml registration with MINIMAX_API_KEY env var
• JSON mode detection for minimax.io base URL in model.py
• API key URL mapping in models router

Frontend (5 files)

• MiniMax provider icon
• Icon registration in icons.ts
• i18n strings for EN, ZH_CN, ZH_TW, JA locales

Documentation (5 files)

• All 4 README files (EN/ZH/ZH_Hant/JA) updated with MiniMax in provider list
• CONFIGURATION_GUIDE.md provider table updated

Tests (2 files, 34 tests)

• 28 unit tests: provider init, model creation, temperature clamping, parameter merging, embedding rejection, factory registration, JSON mode detection, YAML validation, config registration
• 6 integration tests: ConfigManager loading, validation, factory end-to-end model creation, enabled providers list, available models

Configuration

# Set API key
export MINIMAX_API_KEY=your-key-here

# Optionally set as primary provider
export PRIMARY_PROVIDER=minimax

Models

Model	Context	Description
MiniMax-M2.7	204K	Latest flagship model
MiniMax-M2.7-highspeed	204K	Faster variant
MiniMax-M2.5	204K	Previous generation
MiniMax-M2.5-highspeed	204K	Faster variant

Test Plan

• All 34 unit + integration tests pass (pytest valuecell/adapters/models/tests/test_minimax_provider.py)
• Manual test: configure MINIMAX_API_KEY and verify model creation
• Manual test: verify provider appears in Settings UI

20 files changed, 1195 insertions(+), 9 deletions(-)