You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Add release notes for Roo Code versions 2.2 to 3.13, including key features, provider updates, QOL improvements, bug fixes, and miscellaneous improvements. (#195)
This release cycle, prior to the introduction of MCP and major version 2.2, focused on foundational improvements including experimental diff editing, OpenRouter compression, sound effects, and making Roo Cline (now Roo Code) work in all VSCode-compatible editors and side-by-side with the original Cline.
***Auto-Approval for Writes & Commands (v2.1.2):** Initial support for auto-approval of write operations and command execution.
15
+
***Custom Instructions with `.clinerules` (v2.1.2 & v2.1.9):** Support for `.clinerules` custom instructions, with guidance added to settings.
16
+
***Browser Action Auto-Approval (v2.1.3):** Allowed browser actions without approval when `alwaysAllowBrowser` is true.
17
+
18
+
## Compatibility & Core Updates
19
+
20
+
***Runs in All VSCode-Compatible Editors (v2.1.6):** Roo Cline (now Roo Code) was updated to run in all VSCode-compatible editors.
21
+
***Side-by-Side with Cline (v2.1.4):** Enabled Roo Cline (now Roo Code) to run side-by-side with the original Cline extension.
22
+
***Updated Extension Icon and Metadata (v2.1.7).**
23
+
24
+
## QOL Improvements
25
+
26
+
***Larger Text Area & Image Drag-and-Drop (v2.1.21):** Implemented a larger text area input with the ability to drag images into it.
27
+
***Gemini 2.0 Added (v2.1.20).**
28
+
***Copy Prompts from History (v2.1.16).**
29
+
30
+
## Bug Fixes
31
+
32
+
***Sound Effects Settings (v2.1.13):** Fixed issue where sound effects were not respecting settings.
33
+
***Browser Action Approval Fix (v2.1.5).**
34
+
***Diff `write_to_file` Rejection (v2.1.14):** If diffs were enabled, `write_to_file` commands that would lead to truncated output were automatically rejected.
35
+
***Diff Editing Bugfix for Windows Line Endings (v2.1.18).**
36
+
***Improved Error Handling for Diff Editing (v2.1.19).**
This release cycle introduced Model Context Protocol (MCP) support, enabling custom tools, along with numerous settings for finer control over API requests, diff editing, auto-approval, and various QOL improvements and bug fixes.
4
+
5
+
## Key Features & Enhancements
6
+
7
+
***Model Context Protocol (MCP) Support (v2.2.0):**
8
+
* Added support for MCP, enabling Roo Code to use custom tools like web search or GitHub tools.
9
+
* Added an MCP server management tab.
10
+
* Enabled Roo Code to dynamically create new MCP servers based on user requests.
11
+
***API Configuration Profiles (v2.2.45):** Added the ability to save different API configurations to quickly switch between providers and settings. (thanks @samhvw8!)
12
+
***Configurable API Request Retries (v2.2.44):** Allowed automatic retries for failed API requests with a configurable delay. (thanks @RaySinner!)
13
+
***Message Deletion (v2.2.43):** Added functionality to delete single messages or all subsequent messages.
14
+
***Git Context Mentions (v2.2.42):** Added a Git section to the context mentions.
This release cycle marked the introduction of Chat Modes in Roo Code, allowing users to interact with the AI for architectural discussions and codebase questions without immediately diving into code generation. It also included some minor fixes and updates.
4
+
5
+
## Key Feature: Chat Modes (v3.0.0)
6
+
7
+
***Introduction of Chat Modes:** Roo Code now features distinct chat modes (Ask, Architect, Code). This allows users to:
8
+
* Ask questions about system architecture or the codebase.
9
+
* Engage in higher-level discussions before code implementation.
10
+
* Assign different API configuration profiles to each mode, enabling the use of different models for thinking versus coding.
11
+
***Roo Code Reddit Community:** Launched the [Roo Code Reddit](https://www.reddit.com/r/RooCode) for feedback and community interaction.
12
+
13
+
## Bug Fixes & Minor Updates
14
+
15
+
***Reddit Link Fix (v3.0.1):** Corrected the Reddit link in the application.
16
+
***Chat Input Visual Glitch (v3.0.1):** Fixed a small visual glitch in the chat input.
17
+
***Button Alignment Tweaks (v3.0.2):** Made minor adjustments to button alignment in the chat input.
18
+
***VSCode Engine Requirement (v3.0.3):** Updated the required VSCode engine to `^1.84.0` to align with Cline.
This release cycle introduced customizable role definitions and instructions for chat modes, revamped Prompt Enhancements, added DeepSeek-R1 support, an experimental unified diff algorithm, Mistral provider integration, and several bug fixes related to configuration profiles and UI.
4
+
5
+
## Key Features & Enhancements
6
+
7
+
***Customizable Chat Mode Prompts (v3.1.0):** You can now customize the role definition and instructions for each chat mode (Code, Architect, and Ask) via the Prompts tab or mode-specific `.clinerules-mode` files.
8
+
***Revamped Prompt Enhancements (v3.1.0):** The "Enhance Prompt" button now works with any provider and API configuration, allowing for fully customizable prompts.
9
+
***Auto-Approve Chat Bar (v3.1.3):** Added an auto-approve chat bar for quicker interactions. (thanks Cline!)
10
+
***Experimental Unified Diff Algorithm (v3.1.7):** An experimental new unified diff algorithm can be enabled in settings. (thanks @daniel-lxs!)
11
+
12
+
## Provider Updates
13
+
14
+
***DeepSeek-R1 Support (v3.1.7):** Added support for the DeepSeek-R1 model. (thanks @philipnext!)
15
+
***Mistral Provider (v3.1.6):** Integrated the Mistral provider. (thanks Cline!)
16
+
***VS Code Language Models (Experimental) (v3.1.2):** Added experimental support for VS Code Language Models, including Copilot. (thanks @RaySinner / @julesmons!)
17
+
***PKCE Support for Glama (v3.1.2):** Implemented PKCE support for the Glama provider. (thanks @punkpeye!)
18
+
19
+
## QOL Improvements & UI
20
+
21
+
***Copy Markdown from Chat (v3.1.0):** Added a button to copy Markdown content directly from the chat.
22
+
***Fuzzy Search Improvements (v3.1.2):** Enhanced fuzzy search in mentions, history, and model lists. (thanks @samhvw8!)
23
+
***Visual Fixes for Light+ Themes (v3.1.1):** Addressed visual glitches in chat input and settings for light+ themes.
24
+
25
+
## Bug Fixes
26
+
27
+
***Configuration Profile Fixes (v3.1.7, v3.1.6, v3.1.2):** Addressed multiple bugs related to configuration profile saving and switching, particularly for VSCode LM. (thanks @samhvw8!)
28
+
***Auto Approve Menu Fixes (v3.1.4 - v3.1.5):** Corrected bugs in the auto-approve menu.
29
+
***VS Code Language Models Integration Bug (v3.1.3):** Fixed a bug with the VS Code Language Models integration.
30
+
31
+
## Misc Improvements
32
+
33
+
***Updated VSCode Engine Requirement (v3.0.3):** Updated the required VSCode engine to `^1.84.0` to match Cline.
34
+
***Use 'developer' message for o1 system prompt (v3.1.2).**
This release cycle focused on enhancing context management with partial file reads and suggested responses, improved OpenRouter integration, and various fixes for stability and user experience.
4
+
5
+
## Enhanced Context Management
6
+
7
+
***Reading Large Files in Chunks:** Roo Code can now read large files in chunks, allowing for more efficient processing and context utilization. (thanks samhvw8!)
8
+
***Suggested Responses to Questions:** When Roo asks a question, it now provides suggested responses to streamline user interaction. (thanks samhvw8!)
9
+
***More Consistent @-mention Lookups:** Improved lookup consistency for files and folders using @-mentions.
10
+
***Consolidated Code Actions:** Code actions are now grouped into a submenu for a cleaner interface. (thanks samhvw8!)
11
+
12
+
## Provider & API Enhancements
13
+
14
+
***Dynamically Fetched Instructions:** Instructions for creating/editing custom modes and MCP servers are now fetched dynamically. (thanks diarmidmackenzie!)
15
+
***Gemini 2.5 Pro Added:** The Gemini 2.5 Pro model has been added to the Google Gemini provider. (thanks samsilveira!)
16
+
***OpenRouter Token Tracking:** Switched to a more direct method of tracking OpenRouter token usage and spend.
17
+
***Control Over Partial Read Chunk Size:** Users now have more control over the chunk size for partial file reads.
18
+
***R1 Support for OpenAI Compatible:** Added an R1 support checkbox to the OpenAI compatible provider for QWQ. (thanks teddyOOXX!)
19
+
***Bedrock Application Inference Profile:** Added Bedrock support for application-inference-profile. (thanks maekawataiki!)
20
+
***Task Created Event in API:** Added a `taskCreated` event to the API. (thanks wkordalski!)
21
+
22
+
## QOL Improvements
23
+
24
+
***Updated Welcome Page:** The welcome page now provides 1-click OAuth flows with LLM routers. (thanks dtrugman!)
25
+
***New Task Command:** Added a "New Task" command in the Command Palette. (thanks qdaxb!)
26
+
***UX for Chat Text Area:** Updated the user experience for the chat text area. (thanks chadgauth!)
27
+
***Custom Storage Path for Tasks:** Supported a custom storage path for tasks. (thanks Chenjiayuan195!)
28
+
29
+
## Bug Fixes
30
+
31
+
* Fixed MCP error logging (thanks aheizi!).
32
+
* Improved `search_files` tool formatting and logic (thanks KJ7LNW!).
33
+
* Fixed changelog formatting in GitHub Releases (thanks pdecat!).
34
+
* Fixed a bug causing task history to be lost when using WSL.
35
+
* Fixed issues where questions and suggestions weren’t showing for non-streaming models or were hard to read.
36
+
* Fixed opacity of drop-down menus in settings (thanks KJ7LNW!).
37
+
* Fixed bugs with reading and mentioning binary files like PDFs.
38
+
* Fixed pricing information for OpenRouter free models (thanks Jdo300!).
39
+
* Fixed an issue with unit tests on Windows (thanks diarmidmackenzie!).
40
+
* Fixed a maxTokens issue for the Outbound provider (thanks pugazhendhi-m!).
41
+
* Fixed a line number issue with partial file reads (thanks samhvw8!).
42
+
* Fixed context mentions on Windows.
43
+
* Fixed German translations (thanks cannuri!).
44
+
* Fixed telemetry banner internationalization.
45
+
* Corrected max output tokens for Sonnet 3.7 non-thinking models.
46
+
* Fixed logic for firing task completion events.
47
+
* Updated value of max tokens for `gemini-2.5-pro-03-25` (thanks linegel!).
48
+
49
+
## Misc Improvements
50
+
51
+
* Added fake provider for integration tests (thanks franekp!).
52
+
* Reflected Cross-region inference option in ap-xx region (thanks Yoshino-Yukitaro!).
53
+
* Made partial file reads backwards-compatible with custom system prompts.
54
+
* A variety of fixes and improvements to experimental multi-block diff (thanks KJ7LNW!).
55
+
* Renamed and migrated global MCP and modes files (thanks StevenTCramer!).
56
+
* Added `watchPaths` option to McpHub for file change detection (thanks 01Rian!).
57
+
* Enabled reading of image responses from MCP calls (thanks nevermorec!).
0 commit comments