feat: add OpenAI thinking effort setting

[buhagee]

📝 What does this change?

Adds the opencodego.thinking.openai setting for OpenAI GPT models, including support for off, low, medium, high, and xhigh reasoning effort. The setting maps to the Responses API reasoning.effort field.

🧪 How did you test it?

Validated the package configuration and ran the full test suite:

• npm run compile
• npm run lint
• npm test
• npm run package
• npm run test-retry — 7/7 passed

Live testing with an OpenAI model was not performed.

✅ Checklist

• npm run compile passes
• npm run lint passes
• npm test passes
• npm run package produces a VSIX
• I tested it works
• I updated docs/CHANGELOG if needed

[ltmoerdani]

Nice one @buhagee, thanks for this! This closes a gap we missed on our side: the OpenAI thinking strategy, the config key, and the defaults all landed earlier with the per-provider thinking refactor, but the package.json declaration never made it in, so the setting stayed invisible in the Settings UI.

I checked the details as well. The enum values match OPENAI_EFFORTS in src/thinking/openai.ts and the default matches THINKING_DEFAULTS, so it's declaration-only with no behavior change. The payload shape (reasoning: { effort }) is already covered by unit tests, so skipping the live model test is fine here.

One small note, nothing blocking: the settings table in docs/features/02-20260517-per-model-thinking-controls.md doesn't have a row for this setting yet. I'll add that myself after merging, no action needed on your end.

Merging this with a merge commit.