feat: upgrade MiniMax default model to M3

[octo-patch]

Summary

Upgrade the default MiniMax model in the built-in provider list to MiniMax-M3.

Changes

• apps/cli/src/commands/model-support.ts: add MiniMax-M3 as the first entry in the minimax provider list so it becomes the default; remove MiniMax-M2.5 / MiniMax-M2.5-highspeed; keep MiniMax-M2.7 and MiniMax-M2.7-highspeed as alternatives.
• README.md / README.zh-CN.md: refresh the MiniMax row in the supported-models table to match.

Resulting model list

minimax: [
    "MiniMax-M3",
    "MiniMax-M2.7",
    "MiniMax-M2.7-highspeed",
]

Why M3

• MiniMax-M3 is the latest model with a 512K context window, up to 128K output, and image input support, making it the most capable default for Understudy.
• The M2.7 / M2.7-highspeed variants remain available as fallbacks for users who prefer the previous generation or need the low-latency profile.
• The older M2.5 / M2.5-highspeed entries are removed from the built-in registry; they can still be reached via custom model entries if needed.

Validation

• pnpm install (clean)
• pnpm lint — 0 errors (5 pre-existing warnings, unrelated)
• pnpm typecheck — all packages pass
• pnpm build — all packages compile cleanly via tsc -p tsconfig.build.json
• API endpoint / MINIMAX_API_KEY env var / minimax-cn China endpoint are unchanged.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 0e2e6f5ae2

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".