- Modified ToolRepetitionDetector to skip repetition detection for browser_action scroll_down and scroll_up actions
- Added isBrowserScrollAction() helper method to identify scroll actions
- Added comprehensive tests for the new behavior
- Fixes issue where multiple scroll actions were incorrectly flagged as being stuck in a loop
Resolves: https://github.com/RooCodeInc/Roo-Code/discussions/7470
Co-authored-by: Roo Code <roomote@roocode.com>
* fix: change default showRooIgnoredFiles to false to hide ignored files
- Changed default value from true to false across all files
- Updated tests to reflect the new default behavior
- This prevents ignored files from appearing in environment details
Fixes#7368
* fix: update tests to match new showRooIgnoredFiles default
* fix: update test expectation to match new showRooIgnoredFiles default value
The PR changed the default value of showRooIgnoredFiles from true to false,
so the test needs to expect false instead of true when calling formatFilesList.
---------
Co-authored-by: Roo Code <roomote@roocode.com>
Co-authored-by: daniel-lxs <ricciodaniel98@gmail.com>
• Replace dual Set<string> (acceptedFiles/rejectedFiles) with single Map<string, string> (acceptedBaselines) for cleaner state management
• Remove complex hiding/unhiding logic in applyPerFileBaselines()
• Rejected files are simply removed from changeset and reappear naturally when edited again via update FCOAfterEdit
• Accepted files get per-file baselines to show only incremental changes
• Self-correcting system: file visibility determined by diffs, not flags
Problem:
- FCO missing last edited file (calculated at checkpoint creation before tools execute)
- FCO disappears when tasks are aborted (state not preserved)
- Manual user edits must remain protected during rollback (issue #4827)
Solution:
- Add immediate FCO updates after each file editing tool execution
- Preserve FCO state during task abort and restore on resume
- Maintain checkpoint timing BEFORE edits for rollback safety
- Add final checkpoint on task completion to capture all changes
Changes:
- Add updateFCOAfterEdit helper to calculate and display changes without checkpoints
- Update presentAssistantMessage to call FCO updates after file tools
- Add final checkpoint in attemptCompletionTool
- Preserve/restore FCO state in ClineProvider during abort/resume
- Add test utilities for checkpoint functionality
This separates FCO visibility (immediate updates) from checkpoint safety (before edits),
solving both user experience issues while maintaining rollback protection.
- Fix checkpoint memory leak by making
ongoingCheckpointSaves task-scoped
- Moved ongoingCheckpointSaves Map from module-level to
Task class property
- Add cleanup in Task.dispose() method to prevent memory
leaks
- Update checkpoint functions to use task-scoped Map
- Fix test mock to include ongoingCheckpointSaves
property
Applies commits 04bd21403 and 3d15bba1d from backup branch:
FCO Edge Case Fixes:
- Add .roo/ exclusion to checkpoint diffs
- Filter out directories from ShadowCheckpointService.getDiff()
- Implement improved line-by-line diff calculation in FileChangeManager
- Add comprehensive FileChangeManager tests (70+ test cases)
Windows Compatibility:
- Fix 'core.bare and core.worktree do not make sense' error
- Add core.bare=false configuration for shadow git repos
LLM-Only Filtering:
- Complete async integration of getLLMOnlyChanges() in FCO handlers
- Fix getCurrentCline → getCurrentTask method name alignment
- Update all FCO message handlers to use LLM-only filtering
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Change getCurrentCline() to getCurrentTask()
- Fix Promise return type in test mock
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Added more tests for checkpoints
- Fixed bug where fco was not updating after checkpoints properly.
- It would include all previous edits from the previous checkpoint on tasks that were mid edit.
- Added timestamp checking to cover the edge case.
- Added tests for the edge cases covered.
- Bug fix for users enabling FCO after some time using the task.
- One more edge case being covered when the user enables fco after talking with the task for a while.
- To solve this, added test cases and have the settings enabled to set the timestamp if it was previously disabled.
- Added better separation of concerns for testing of FCO web ui.
* feat: update /init command to focus only on non-obvious discoveries
- Emphasize that only non-obvious information discovered by reading files should be included
- Update all mode-specific examples to show truly non-obvious rules
- Add CRITICAL section explaining what qualifies as non-obvious
- Exclude standard patterns and framework defaults
- Focus on gotchas, hidden requirements, and counterintuitive patterns
* fix: ensure /init command checks for and updates existing AGENTS.md files
- Check for existing AGENTS.md in root directory
- Check for existing AGENTS.md in all .roo/rules-*/ directories
- Read and improve existing files rather than replacing them
- Update discovery phase to explicitly look for these files
* refactor: clean up /init command to properly handle existing AGENTS.md files
- Consolidate all AGENTS.md path checking into step 1 with CRITICAL emphasis
- Add explicit instructions for deep dive improvement of existing files
- Remove redundant mentions throughout the workflow
- Emphasize iterating and enhancing existing files, not just appending
- Add clear guidance on cleaning up outdated info and reorganizing
- Make discovery phase the single source of truth for file checking
* fix: strengthen /init command to aggressively remove obvious information
- Add explicit instructions to DELETE obvious information first
- Emphasize that files should get SHORTER, not longer
- Use stronger language: CRITICALLY EVALUATE, AGGRESSIVELY DELETE
- Make clear that standard practices must be removed even if previously included
- Add success metric: files should be more concise and valuable
- Goal is to fix the tendency to just append instead of clean up
* fix: ensure .roo directory is created in project root, not system root
CRITICAL FIX: Paths were being interpreted as absolute, causing .roo to be created at system root
- Clarify all paths are relative to PROJECT/WORKSPACE root
- Add explicit warnings that .roo must be in project root, not system root
- Update example to show proper project structure
- Add IMPORTANT notes throughout emphasizing relative paths
- Fix prevents accidental creation of /.roo at system level
* test: update built-in-commands test to match new init command content
- Replace 'mode-specific rule directories' with 'non-obvious'
- Replace 'analysis_workflow' with 'discovered by reading files'
- Tests now align with the new focus on non-obvious discoveries
* fix: resolve newTaskRequireTodos setting not working correctly
- Use dynamic Package.name instead of hardcoded namespace values
- Show todos parameter as optional/required based on setting value
- Remove hardcoded new_task example from shared tool use section
- Update tests to use Package.name pattern
The setting now works correctly for both regular and nightly builds
without requiring hardcoded namespace values.
* test: update snapshots after removing hardcoded new_task example
The snapshots needed updating because the hardcoded new_task example
was removed from the shared tool use section
* fix: remove namespace prefix from newTaskRequireTodos setting
The setting name should not include the namespace prefix in package.json
as VSCode automatically adds the extension's namespace. This was preventing
the setting from appearing in the VSCode settings UI.
* fix: restore prefixed setting id roo-cline.newTaskRequireTodos for Settings UI visibility
* docs(newTaskTool): explain dynamic Package.name config namespace\n\ntest(newTaskTool): verify config uses Package.name variant (roo-code-nightly)
* refactor: simplify new-task prompt generation
- Replace complex template literals with two complete prompt constants
- Remove nested ternary operators for better readability
- Hide todos parameter completely when disabled (not shown as optional)
- Update tests to reflect new behavior
- Reduce code from 105 to 66 lines for better maintainability
* test: update snapshots after removing optional todos parameter
The todos parameter is now conditionally required based on the newTaskRequireTodos setting, so the snapshots needed to be updated to reflect the new tool documentation format.
---------
Co-authored-by: daniel-lxs <ricciodaniel98@gmail.com>
* fix: handle Mistral thinking content as reasoning chunks
- Add TypeScript interfaces for Mistral content types (text and thinking)
- Update createMessage to yield reasoning chunks for thinking content
- Update completePrompt to filter out thinking content in non-streaming mode
- Add comprehensive tests for reasoning content handling
- Follow the pattern used by other providers (Anthropic, OpenAI, Gemini, etc.)
Fixes#6842
* fix: resolve TypeScript type issue in completePrompt method
* fix: handle Mistral thinking content chunks in streaming responses
- Added ContentChunkWithThinking type helper to handle thinking chunks
- Properly converts thinking content to reasoning chunks in streaming
- Filters out thinking content in non-streaming completePrompt responses
- Confirmed that Mistral API does send thinking chunks with type 'thinking'
- Works with Mistral SDK v1.9.18
---------
Co-authored-by: Roo Code <roomote@roocode.com>
Co-authored-by: daniel-lxs <ricciodaniel98@gmail.com>
