feat: add MiniMax as LLM provider

[octo-patch]

Summary

Add @llm-tools/embedjs-minimax package providing MiniMax model support via OpenAI-compatible API.

What's included

• MiniMax LLM model class extending BaseModel, using @langchain/openai ChatOpenAI with MiniMax base URL
• Models supported: MiniMax-M2.7 (default, 204K context) and MiniMax-M2.7-highspeed (204K context)
• Temperature clamping to MiniMax's valid range (0.0, 1.0]
• Think-tag stripping — automatically removes <think>...</think> reasoning blocks from M2.7 responses
• Token usage reporting via usage_metadata
• Env var support — reads MINIMAX_API_KEY when apiKey not passed
• 14 tests (4 integration + 6 temperature + 4 constructor)

Usage

import { RAGApplicationBuilder } from '@llm-tools/embedjs';
import { MiniMax } from '@llm-tools/embedjs-minimax';
import { OpenAiEmbeddings } from '@llm-tools/embedjs-openai';

const app = await new RAGApplicationBuilder()
    .setModel(new MiniMax({ modelName: 'MiniMax-M2.7' }))
    .setEmbeddingModel(new OpenAiEmbeddings())
    .build();

Files changed

File	Description
models/embedjs-minimax/src/minimax-model.ts	MiniMax model class
models/embedjs-minimax/src/index.ts	Package exports
models/embedjs-minimax/package.json	Package manifest
models/embedjs-minimax/project.json	Nx project config
models/embedjs-minimax/tsconfig.json	TypeScript config
models/embedjs-minimax/eslint.config.js	ESLint config
models/embedjs-minimax/README.md	Usage documentation
models/embedjs-minimax/tests/integration.test.mjs	Tests
tsconfig.base.json	Added path alias

Test plan

• All 14 tests pass (node --test models/embedjs-minimax/tests/integration.test.mjs)
• npx nx build embedjs-minimax compiles successfully
• Verified API responses from both M2.7 and M2.7-highspeed models
• Verified token usage metadata is returned
• Verified temperature clamping behavior

[sonarqubecloud]

Quality Gate passed

Issues
17 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud