* feat: add markdown support to codebase indexing (#4660)
* fix: implement chunking for large markdown sections and fix Qdrant deduplication issue (#4660)
- Modified parseMarkdownContent to chunk large sections (>1150 chars)
- Added support for chunking header-less markdown files
- Fixed _chunkTextByLines to handle oversized lines properly
- Added defensive check for parseMarkdown returning undefined
- Fixed Qdrant ID generation to use segmentHash instead of file:line
- This was the root cause: chunks were being deduplicated
- Each chunk now gets a unique ID even from the same line
- Added comprehensive tests for all edge cases
- Ensures all markdown content is properly indexed in Qdrant
* fix: remove redundant supported-extensions test file
As identified in PR review, the supported-extensions.spec.ts file only tests
the contents of an array, which is already implicitly covered by the functional
tests in parser.spec.ts. Removing this reduces maintenance overhead without
sacrificing test quality.
* test: address PR review feedback
- Remove redundant test 'should handle large markdown documentation folders efficiently'
that only verified the scanner could iterate over mocked files
- Add test to verify unique point IDs are generated for each block from the same file,
ensuring the segmentHash-based ID generation prevents collisions
* fix: add segmentHash to vector point payload and refactor tests
- Add segmentHash to payload in scanner.ts to fix vector point ID generation
- Split parser.spec.ts tests into focused unit tests (mocked dependencies)
- Move integration tests to new markdownIntegration.spec.ts file
- Each test suite now has clear, distinct responsibilities
- Fixes issue #4660: vector point ID collisions for large Markdown files
* refactor: move redundant tests
* feat: enhance markdown processing with consistent chunking logic and segment hashing
---------
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
* fix(tools): Resolve intermittent hangs and lack of clear error feedback in apply_diff tool (#5390)
* fix: show user-friendly error message for XML parsing failures
- Keep detailed XML structure instructions only for LLM via pushToolResult
- Show simple error message to user via cline.say
- Aligns with established error handling pattern in codebase
* feat: add telemetry for XML parsing errors
- Track XML parsing errors using captureDiffApplicationError
- Include consecutive mistake count for better error analysis
- Helps monitor and debug XML parsing issues in production
---------
Co-authored-by: Example User <user@example.com>
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
* fix: prevent chatbox focus loss during automated file editing (#4574)
- Add preserveFocus: true to DiffViewProvider openDiffEditor method
- Implement comprehensive focus preservation during cursor positioning and scrolling
- Make scrollToFirstDiff async with focus restoration capabilities
- Update all tool files to use async scrollToFirstDiff
- Add comprehensive unit and E2E tests for focus preservation
Fixes#4574
* refactor: extract focus restoration logic and improve consistency
- Created restoreEditorFocus() helper method to eliminate code duplication
- Changed preserveFocus from false to true for consistent behavior
- Added error handling with silent logging for focus restoration failures
- Added null checks for undefined activeTextEditor cases
* Delete apps/vscode-e2e/src/suite/tools/focus-preservation.test.ts
It doesn't seem to be testing anything
* refactor: remove focus restoration logic to simplify diff editor interactions
* refactor: remove unnecessary await from scrollToFirstDiff calls
---------
Co-authored-by: Daniel <57051444+daniel-lxs@users.noreply.github.com>
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
* Fix code index secret persistence with async VSCode storage
- Add async secret methods to CodeIndexConfigManager
- Implement direct VSCode secret storage access bypassing ContextProxy cache
- Update loadConfiguration to use async secret loading
- Modify webview message handler to use new async secret storage
- Add public secret methods to CodeIndexManager
- Enhance debugging throughout secret flow
This fixes the issue where API keys were saved but not loaded immediately
into services due to ContextProxy cache synchronization issues.
* Fix code index secret persistence and test failures
- Add async secret handling to CodeIndexConfigManager with new methods:
- getSecretAsync(), storeSecretAsync() for individual secrets
- loadSecretsAsync(), storeSecretsAsync() for batch operations
- Update doesConfigChangeRequireRestart() to check OpenAI Compatible modelDimension changes
- Fix all failing tests by using setupSecretMocks() helper consistently
- Update manager.spec.ts to properly mock _recreateServices to avoid real service creation
This ensures API keys and other secrets are properly loaded from VSCode's async secret storage
and that configuration changes requiring service restart are correctly detected.
* feat: improve code index settings secret handling in UI
- Show placeholder dots (••••••••••••••••) in password fields when secrets are already set
- Only send modified secret fields to prevent overwriting existing secrets with empty values
- Track which fields have been modified by the user
- Add requestCodeIndexSecretStatus message handler to check if secrets exist
- Fix console.log to handle empty string keys without errors
- Ensure changing one setting doesn't clear other unmodified secrets
* refactor: disconnect code index from unified settings system
- Rename handleExternalSettingsChange to handleSettingsChange for clarity
- Remove handleSettingsChange call from ClineProvider (not related to code index)
- Remove codebaseIndexConfig from general settings save in SettingsView
- Delete unused codebaseIndexConfig message handler
- Remove codebaseIndexConfig from WebviewMessage type definition
- Code index settings are now fully independent with their own dedicated UI
* feat: separate code index enable/disable from indexing settings
- Move 'Enable codebase indexing' toggle to global settings in Experimental section
- Keep indexing-specific settings (API keys, URLs, models) in dedicated Code Index Settings component
- Add codebaseIndexEnabled handler to webview message handler
- Update translations with new settings title and disabled message
- Ensure code index service properly responds to enable/disable changes
- Maintain backward compatibility with existing codebaseIndexConfig structure
* refactor: remove ContextProxy.getVSCodeContext() and pass ExtensionContext directly
- Updated CodeIndexConfigManager to accept vscode.ExtensionContext in constructor
- Modified CodeIndexManager to pass context directly to CodeIndexConfigManager
- Updated webviewMessageHandler to use provider.context.secrets directly
- Removed getVSCodeContext() method from ContextProxy
- Updated all related tests to reflect these changes
- Fixed CodeIndexSettings webview tests after UI changes
* refactor: streamline secret handling by removing async methods and utilizing ContextProxy directly
* feat: translations and popover component
* refactor: simplify test mocks and improve checkbox handling in CodeIndexSettings tests
* refactor: remove debug logging from CodeIndexConfigManager, CodeIndexManager, CodeIndexServiceFactory, and QdrantVectorStore
* fix: merge missing translation keys from main after rebase
- Add advancedConfigLabel, searchMinScoreLabel, searchMinScoreDescription, searchMinScoreResetTooltip keys
- Update startIndexingButton and clearIndexDataButton labels to match main
- Preserve all CodeIndexPopover translations added in this PR
* Revert "fix: merge missing translation keys from main after rebase"
This reverts commit beb1de4924ac1475731fcd06d994ddb96eb1e5fd.
* fix: add missing translation keys from main branch after rebase
- Added codeIndex.advancedConfigLabel
- Added codeIndex.searchMinScoreLabel
- Added codeIndex.searchMinScoreDescription
- Added codeIndex.searchMinScoreResetTooltip
These keys exist on main but were missing from non-English locales after rebase.
* fix: remove clickIndicatorMessage, fix toggle message type, and clean up translation key inconsistencies
* refactor: streamline settings management in CodeIndexPopover and improve secret handling
* refactor: remove debug logging from configuration checks in CodeIndexConfigManager and webviewMessageHandler
* fix: translations
* refactor: remove CodeIndexSettings component and associated tests
* feat: Add folder deletion when custom mode is deleted (#5210)
* fix: address PR feedback - fix Korean translation typo and add missing German translation
* fix: use os.homedir() instead of vscode.env.userHome
* fix: use dynamic home directory in webview tests for cross-platform compatibility
* fix: normalize path separators in tests for Windows compatibility
* feat: implement DeleteModeDialog component for mode deletion confirmation
* fix: use ref to store current modeToDelete value for consistent state handling
* feat: add error handling for rules folder deletion and localize error messages
* Update src/i18n/locales/ja/common.json
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
* Update src/i18n/locales/zh-CN/common.json
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
* fix: remove redundant rulesFolderExists from deleteCustomModeCheck response
* fix: update webviewMessageHandler tests to match fs import style
---------
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
Co-authored-by: Daniel <57051444+daniel-lxs@users.noreply.github.com>
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
Co-authored-by: Daniel <57051444+daniel-lxs@users.noreply.github.com>
Co-authored-by: Matt Rubens <mrubens@users.noreply.github.com>
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
Co-authored-by: Daniel <57051444+daniel-lxs@users.noreply.github.com>
* Fixes#5294: Add instruction to Architect mode to avoid time estimates
- Updated Architect mode custom instructions to explicitly prohibit providing time estimates
- Added clear guidance to focus on actionable plans without speculating about implementation timeframes
- Updated test snapshot to reflect the new instruction
- All tests passing
* More tweaks to prompt
---------
Co-authored-by: Matt Rubens <mrubens@users.noreply.github.com>
* chore: adding x-title header and testing for litellm
* chore: indentation fi and headers order fix
* chore: spacing fix
* chore: removed white space
* fix: allow user headers to override default headers and clean up formatting
- Reorder header spread in router-provider.ts so user-provided openAiHeaders can override DEFAULT_HEADERS
- Remove unnecessary blank lines after imports for consistency
- This matches the pattern used in openai.ts where DEFAULT_HEADERS come first
---------
Co-authored-by: Brendan-Z <brendanzhou.99@gmail.com>
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
* feat: change default mode to architect for new installations
- Reordered modes array to put architect mode first
- Updated test to reflect architect as the default mode
- Maintains backward compatibility for existing users
* test: update snapshots and fix tests for architect default mode
- Update all snapshot tests to reflect architect as the default mode
- Fix mode destructuring in validateToolUse tests
- All tests now pass with the new default mode configuration
* docs: add comment explaining mode ordering for default selection
Addresses PR review feedback to clarify that the first mode in the array
serves as the default for new installations
* fix: Update integration test to use lowercase mode name
The test was using 'Ask' (capitalized) but modes are now lowercase 'ask'.
This was causing the test to timeout as the mode validation failed.
* https parsing in URL sanatizing for git repo
* fix: remove duplicate JSDoc and add SSH to HTTPS conversion test
- Remove duplicate JSDoc comment from convertGitUrlToHttps function
- Add test case to verify getGitRepositoryInfo converts SSH URLs to HTTPS format
---------
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
* fix: resolve E2BIG error by passing large prompts via stdin to Claude CLI
- Pass messages via stdin instead of command line arguments to avoid Linux argument length limits
- Add --input-format text flag to claude CLI command
- Update execa configuration to use stdin pipe
- Fix corresponding unit tests with proper async iterator mocking
- Resolves spawn E2BIG errors when using very large conversation histories
* fix: address race condition and improve error handling
- Use setImmediate to ensure process is spawned before writing to stdin
- Add proper error handling for stdin write operations
- Add tests for error scenarios
- Update existing tests to handle async behavior properly
* fix: remove --input-format text flag to prevent CLI parsing errors
The --input-format text flag was causing the Claude CLI to misinterpret
the JSON content passed via stdin, leading to errors like 'unknown option -------'
when the system prompt contained dashes. Removing this flag allows the CLI
to properly handle the JSON input via stdin.
---------
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
## Fix terminal keyboard shortcut error when adding content to context
**Fixes:** [#2276](https://github.com/RooCodeInc/Roo-Code/issues/2276)
**Problem:**
- Using keyboard shortcuts to add terminal content to context threw error: "Cannot read properties of undefined (reading 'selection')"
- Context menu worked correctly, but keyboard shortcuts failed
**Root Cause:**
- Command handler accessed `args.selection` without null checking
- When triggered via keyboard shortcut, VS Code passes `undefined` for `args` parameter
- When triggered via context menu, VS Code passes an object with `selection` property
**Solution:**
- Changed `args.selection` to `args?.selection` using optional chaining
- Maintains existing fallback behavior when no selection is available
- Preserves backward compatibility with context menu functionality
**Files Modified:**
- `src/activate/registerTerminalActions.ts` - Added null safety for args parameter
**Testing:**
- ✅ Keyboard shortcuts now work without errors
- ✅ Context menu functionality preserved
- ✅ Fallback to `Terminal.getTerminalContents()` works in both scenarios
Co-authored-by: ellipsis-dev[bot] <65095814+ellipsis-dev[bot]@users.noreply.github.com>
Co-authored-by: Matt Rubens <mrubens@users.noreply.github.com>
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
Co-authored-by: Daniel <57051444+daniel-lxs@users.noreply.github.com>
* fix: resolve Claude Code token counting inefficiency and enable caching (#5104)
- Remove 1.5x fudge factor from Claude Code token counting
- Enable prompt caching support for all Claude Code models
- Add comprehensive tests for token counting and caching
- Update existing tests to reflect accurate token counting
This fixes the extreme token inefficiency where simple messages would
jump from ~40k to over 60k tokens, causing API hangs when approaching
the artificial 120k limit. Claude Code now properly utilizes its full
200k context window with accurate token counting.
* fix: address PR review comments
- Extract IMAGE_TOKEN_ESTIMATE as a named constant for clarity
- Update token counting tests to use exact counts instead of ranges for deterministic testing
- Fix test expectations to match actual tokenizer output
* Remove token counting changes, keep only cache support
- Removed custom countTokens override from claude-code.ts
- Deleted claude-code-token-counting.spec.ts test file
- Kept cache token collection and reporting functionality
- Kept supportsPromptCache: true for all Claude Code models
- Kept claude-code-caching.spec.ts tests
This focuses the PR on enabling cache support without modifying token counting behavior.
* fix: update webview test to expect supportsPromptCache=true for Claude Code models
---------
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
* Fix temperature parameter error for Azure OpenAI reasoning models
* Fix tests: Update O3 family model tests to expect temperature: undefined
- Updated failing tests in openai.spec.ts to expect temperature: undefined for O3 models
- This aligns with the PR changes that remove temperature parameter for Azure OpenAI o1, o3, and o4 models
- All 4 previously failing tests now pass
---------
Co-authored-by: Daniel Riccio <ricciodaniel98@gmail.com>
