feat: add MiniMax as first-class LLM provider

[octo-patch]

Summary

Add MiniMax as a built-in LLM service provider for Magic. MiniMax offers high-performance AI models (M2.7, M2.5 series) with million-token context windows, tool calling, and deep thinking capabilities, all served via an OpenAI-compatible API at https://api.minimax.io/v1.

Changes

Backend (PHP)

• ProviderCode.php: Register MiniMax enum case with OpenAIModel implementation
• ProviderTemplateId.php: Add MiniMaxLlm = '23' template ID for provider-category mapping
• ServiceProviderInitializer.php: Add MiniMax provider initialization data with bilingual (EN/CN) descriptions
• LLMMiniMaxProvider.php: Connectivity test class following existing DeepSeek pattern

Frontend (TypeScript)

• aiModel.ts: Add MiniMax to ServiceProvider enum with default API URL

Tests

• ProviderCodeMiniMaxTest.php: 12 unit tests for enum, implementation, sort order, template ID mapping
• LLMMiniMaxProviderTest.php: 5 unit tests for connectivity test with mock HTTP (success, auth error, network error)
• ServiceProviderInitializerMiniMaxTest.php: 4 unit tests for provider data, translations, sort order uniqueness

Integration Notes

• MiniMax uses the standard OpenAI-compatible API format, so it naturally falls through to the default case in ProviderConfigFactory and getImplementationConfig() - no special adapter needed
• Temperature=0 is accepted by the MiniMax API
• Available models: MiniMax-M2.7, MiniMax-M2.7-highspeed, MiniMax-M2.5, MiniMax-M2.5-highspeed

Test Plan

• Unit tests for ProviderCode enum (implementation, sort order, template mapping)
• Unit tests for LLMMiniMaxProvider connectivity test (mock HTTP)
• Unit tests for ServiceProviderInitializer data integrity
• Manual: Configure MiniMax provider in admin UI with API key
• Manual: Verify chat completion through MiniMax models

---

Note

Medium Risk
Adds a new third-party LLM provider option end-to-end (backend enums/templates/seed data plus connectivity check and frontend defaults), which could affect provider selection/order and configuration flows. Risk is mitigated by being largely additive and covered by new unit tests.

Overview
Adds MiniMax as a first-class llm service provider.

Backend registers ProviderCode::MiniMax (mapped to the existing OpenAIModel implementation), introduces a new ProviderTemplateId::MiniMaxLlm, seeds MiniMax into default provider initialization data (with EN/CN translations), and adds an LLMMiniMaxProvider connectivity test that validates the API key by calling GET https://api.minimax.io/v1/models.

Frontend updates AiModel.ServiceProvider and ServiceProviderUrl to include MiniMax with the default base URL, and the PR adds unit tests covering the new enum/template mappings, seeded provider data, and connectivity test behavior.

^{Written by Cursor Bugbot for commit 04f6291. This will update automatically on new commits. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

[cursor]

New connectivity test class duplicates existing DeepSeek provider

Low Severity

LLMMiniMaxProvider is an exact duplicate of LLMDeepSeekProvider — the only difference is the $apiBase URL string. All methods (connectivityTestByModel, fetchModels) have identical logic. Additionally, a grep for both class names shows neither is referenced anywhere in application code outside their own definitions and test files, suggesting these connectivity test classes may be unused dead code. A single parameterized class (or a base class accepting the API URL) would eliminate the duplication.