Implement MiniMax chat completions adapter

[Photon101]

Summary

• replace the MiniMax stub with a real OpenAI-compatible chat completions adapter
• use the documented MiniMax OpenAI-compatible base URL and current M2 model IDs
• support dry-run behavior, standard generation options, opts.extra passthrough for MiniMax options such as reasoning_split, token usage mapping, and concise HTTP error excerpts
• add focused tests for dry-run behavior, request payload/auth, token usage mapping, and error handling

References

• https://platform.minimax.io/docs/api-reference/text-openai-api
• https://platform.minimax.io/docs/api-reference/api-overview

Validation

• corepack pnpm exec vitest run packages/ai/minimax/src/index.test.ts
• corepack pnpm exec tsc -p packages/ai/minimax/tsconfig.json --noEmit
• git diff --check