mirror of
https://github.com/RooVetGit/Roo-Code.git
synced 2026-09-19 00:01:19 +00:00
a1bedb3434
- Add useAdaptiveThinking and adaptiveThinkingEffort settings to provider config
- Add supportsAdaptiveThinking and supportsAdaptiveThinkingMaxEffort model flags
- Implement getAnthropicReasoning() with {type: 'adaptive'} mode
- Implement getAnthropicOutputConfig() with effort parameter (low/medium/high/max)
- Max effort only on Opus 4.6; falls back to high on Sonnet 4.6
- Add adaptive thinking checkbox and effort selector to ThinkingBudget UI
- Fix: Max Output Tokens slider always visible (independent of adaptive thinking)
- Add i18n translations for all 17 supported locales
- Add comprehensive test coverage for reasoning transforms
Fixes #11732
|
||
|---|---|---|
| .. | ||
| __tests__ | ||
| cache-strategy | ||
| caching | ||
| ai-sdk.ts | ||
| anthropic-filter.ts | ||
| bedrock-converse-format.ts | ||
| gemini-format.ts | ||
| image-cleaning.ts | ||
| minimax-format.ts | ||
| mistral-format.ts | ||
| model-params.ts | ||
| openai-format.ts | ||
| r1-format.ts | ||
| reasoning.ts | ||
| stream.ts | ||
| vscode-lm-format.ts | ||
| zai-format.ts | ||