{"version":"2.0.0","anchor":"extended-thinking-simplification","canonical_anchor":"extended-thinking-simplification","heading":"Extended Thinking Simplification","tier":null,"area":null,"scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.0.0\/e\/extended-thinking-simplification","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.0.0","markdown":"### Extended Thinking Simplification\n\n**What:** Streamlined from multilingual multi-level system to single trigger\n\n**Previous (v1.0.128):**\n\n- 8 languages supported (English, Japanese, Chinese, Spanish, French, German, Korean, Italian)\n- 4 thinking levels: HIGHEST (31,999 tokens), MIDDLE (10,000), BASIC (4,000), NONE (0)\n- Dozens of trigger phrases per language\n- Complex pattern matching with negation filtering\n\n**Now (v2.0.0):**\n\n- English-only \"ultrathink\" trigger\n- 2 levels: ULTRATHINK (31,999 tokens) or NONE (0)\n- Simple case-insensitive regex matching\n- New `thinkingMetadata` API for programmatic control\n\n**Implications:**\n\n- More predictable behavior\n- Easier to remember single trigger word\n- Programmatic control via metadata for advanced use cases\n- **Evidence**: Removed `v19` patterns at line 363665 in v1.0.128, new simplified system with `NU1` at line 363628 and functions at lines 363644-363671 in v2.0.0"}