|
| 1 | +--- |
| 2 | +description: Roo Code 3.36.0 introduces non-destructive context management for reliable checkpoint restoration, plus Roo provider enhancements and provider improvements. |
| 3 | +keywords: |
| 4 | + - roo code 3.36.0 |
| 5 | + - context condensing |
| 6 | + - checkpoint restoration |
| 7 | + - rewind |
| 8 | + - roo provider |
| 9 | + - reasoning details |
| 10 | +image: /img/v3.36.0/v3.36.0.png |
| 11 | +--- |
| 12 | + |
| 13 | +# Roo Code 3.36.0 Release Notes (2025-12-04) |
| 14 | + |
| 15 | +This release introduces non-destructive context management for reliable checkpoint restoration, adds reasoning visibility for supported models, and improves provider reliability across Roo, Minimax, Cerebras, and Vercel AI Gateway. |
| 16 | + |
| 17 | +<img src="/img/v3.36.0/v3.36.0.png" alt="Roo Code v3.36.0 Release" width="600" /> |
| 18 | + |
| 19 | +## Non-Destructive Context Management |
| 20 | + |
| 21 | +Context condensing and sliding window truncation now preserve your original messages internally rather than deleting them ([#9665](https://github.com/RooCodeInc/Roo-Code/pull/9665)). When you rewind to an earlier checkpoint, the full conversation history is restored automatically. This applies to both automatic condensing and sliding window operations. |
| 22 | + |
| 23 | +## Provider Updates |
| 24 | + |
| 25 | +* **Reasoning Details Support**: The Roo provider now displays reasoning details from models with extended thinking capabilities, giving you visibility into how the model approaches your requests ([#9796](https://github.com/RooCodeInc/Roo-Code/pull/9796)) |
| 26 | + |
| 27 | +* **Native Tools Default**: All Roo provider models now default to native tool protocol for improved reliability and performance ([#9811](https://github.com/RooCodeInc/Roo-Code/pull/9811)) |
| 28 | + |
| 29 | +* **Minimax search_and_replace**: The Minimax M2 model now uses search_and_replace for more reliable file editing operations ([#9780](https://github.com/RooCodeInc/Roo-Code/pull/9780)) |
| 30 | + |
| 31 | +* **Cerebras Token Optimization**: Conservative 8K token limits prevent premature rate limiting (thanks sebastiand-cerebras!), plus deprecated model cleanup ([#9804](https://github.com/RooCodeInc/Roo-Code/pull/9804)) |
| 32 | + |
| 33 | +* **Vercel AI Gateway**: More reliable model fetching for models without complete pricing information ([#9791](https://github.com/RooCodeInc/Roo-Code/pull/9791)) |
| 34 | + |
| 35 | +## Bug Fixes |
| 36 | + |
| 37 | +* **Write Tool Validation**: Resolved false positives where `write_to_file` incorrectly rejected complete markdown files containing inline code comments like `# NEW:` or `// Step 1:` (thanks Michaelzag!) ([#9787](https://github.com/RooCodeInc/Roo-Code/pull/9787)) |
| 38 | + |
| 39 | +* **Download Count Display**: Fixed homepage download count to display with proper precision for million-scale numbers ([#9807](https://github.com/RooCodeInc/Roo-Code/pull/9807)) |
| 40 | + |
| 41 | +## Misc Improvements |
| 42 | + |
| 43 | +* **Tool Consolidation**: Removed the deprecated `insert_content` tool; use `apply_diff` or `write_to_file` for file modifications ([#9751](https://github.com/RooCodeInc/Roo-Code/pull/9751)) |
| 44 | + |
| 45 | +* **Experimental Settings**: Temporarily disabled the parallel tool calls experiment while improvements are in progress ([#9798](https://github.com/RooCodeInc/Roo-Code/pull/9798)) |
| 46 | + |
| 47 | +* **Infrastructure**: Updated Next.js dependencies for web applications ([#9799](https://github.com/RooCodeInc/Roo-Code/pull/9799)) |
0 commit comments