fix(deepseek): skip model selection when --model is omitted

[jiang1997]

Description

DeepSeek no longer always exposes a model selector. deepseek ask "hello" --new still tried to select the default instant model, so it failed with Could not switch to instant model before sending the prompt.

Only run model selection when --model was explicitly supplied. A default invocation now uses DeepSeek's current model and proceeds to send. Explicit model requests still fail if DeepSeek cannot confirm the requested selection. The adapter documentation now describes this behavior and removes an example that combined incompatible model and search options.

Related issue: #2484.

Type of Change

• 🐛 Bug fix
• ✨ New feature
• 🌐 New site adapter
• 📝 Documentation
• ♻️ Refactor
• 🔧 CI / build / tooling

Checklist

• I ran the checks relevant to this PR
• I updated tests or docs if needed
• I included output or screenshots when useful

Documentation (if adding/modifying an adapter)

• Added doc page under docs/adapters/ (if new adapter)
• Updated docs/adapters/index.md table (if new adapter)
• Updated sidebar in docs/.vitepress/config.mts (if new adapter)
• Updated README.md / README.zh-CN.md when command discoverability changed
• Used positional args for the command's primary subject unless a named flag is clearly better
• Normalized expected adapter failures to CliError subclasses instead of raw Error

Verification

• npm exec vitest run -- --project adapter clis/deepseek/ask.test.js clis/deepseek/utils.test.js — 41 tests passed
• npm run typecheck — passed
• git diff --check origin/main...HEAD — passed

The regression test covers ask --new when the model selector is absent and verifies that sending continues. A separate test keeps the explicit --model instant failure behavior.