changeset

* added posthog remote config

* changeset

* add feature flags constant

* move init outside function to keep from potentially re-running

---------

Co-authored-by: Elephant Lumps <celestial_vault@mac.mynetworksettings.com>

.changeset/blue-lobsters-sort.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+create clinerules folder if its currently a file and creating new rule

.changeset/bright-knives-wait.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+updated drag and drop text to say "drop" instead of "drag"

.changeset/bright-zebras-sip.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Add markdown copy to chat

.changeset/brown-radios-do.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+disabling shift hint for now & improving tooltip behavior

.changeset/clean-needles-deliver.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Add support for custom model ID in AWS Bedrock provider, enabling use of Application Inference Profile.

.changeset/curly-guests-raise.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Full support for LaTeX rendering

.changeset/custom-timeout-setting.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Add support for custom API request timeout. Previously, timeouts were hardcoded to 30 seconds for providers like Ollama or 15 seconds for OpenRouter and Cline. Now users can set a custom timeout value in milliseconds through the settings interface.

.changeset/fair-papayas-learn.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Minor UX improvement to drag and drop ux

.changeset/fresh-hats-dance.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Remove linear pull request action

.changeset/funny-numbers-trade.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+searchCommits protobus migration

.changeset/hungry-islands-jog.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+createRuleFile protobus migration

.changeset/itchy-ligers-give.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Fix Handle @withRetry() SyntaxError when running extension locally issue

.changeset/kind-pears-invent.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Add dependsOn to more blocks in the tasks.json

.changeset/large-crews-divide.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Fix for git commit mentions in repos with no git commits

.changeset/lemon-pears-talk.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+enabled pricing calculation for gemini and vertex + more robust caching & cache tracking for gemini & vertex

.changeset/long-cheetahs-love.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Adding args to allow Cursor to open workspaces (for checkpoint testing/development)

.changeset/loud-spies-deny.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+add truncation notice when truncating manually

.changeset/many-beans-yell.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+feat(bedrock): Introduce Amazon Nova Premier

.changeset/mean-nails-teach.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Migrate Cline Tools Section to new docs

.changeset/nervous-dolphins-greet.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Add a timeout setting for the terminal connection, allowing users to adjust this if they are having timeout issues

.changeset/nervous-eels-look.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Added copy button to code blocks.

.changeset/new-yaks-eat.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+addressed race condition in terminal command usage

.changeset/nine-rings-shout.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+add checkpoints after more messages

.changeset/ninety-numbers-decide.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Introduce UI library for future UI development

.changeset/rare-ladybugs-talk.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+add newrule slash command

.changeset/real-cobras-rule.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Add remote config with posthog allowing for disabling new features until they're reading, making for a better developer experience.

.changeset/rude-ligers-guess.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+checkIsImageURL migrated to protobus

.changeset/serious-baboons-train.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+protobus migration for openImage

.changeset/shy-buckets-itch.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+add cache ui for open router and cline provider

.changeset/six-hairs-rush.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+showing expanded task by default

.changeset/sweet-toes-bake.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Refactor to not pass a message for showing the MCP View from the servers modal

.changeset/tall-cups-fly.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Fix Windows path issue: Correct handling of import.meta.url to avoid leading slash in pathname

.changeset/tall-ghosts-help.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": minor
+---
+
+Migrate the addRemoteServer to protobus

.changeset/thick-rice-rush.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Lowering Gemini cache TTL time

.changeset/thirty-feet-rule.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+Adding UI to show openrouter balance next to provider

.changeset/tough-kangaroos-beam.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+deleteRuleFile protobus migration

.changeset/twenty-comics-enjoy.md

@@ -0,0 +1,5 @@
+---
+"claude-dev": patch
+---
+
+fix cost calculation

.clinerules/cline-overview.md

@@ -0,0 +1,763 @@
+# Cline Extension Architecture & Development Guide
+
+## Project Overview
+
+Cline is a VSCode extension that provides AI assistance through a combination of a core extension backend and a React-based webview frontend. The extension is built with TypeScript and follows a modular architecture pattern.
+
+## Architecture Overview
+
+```mermaid
+graph TB
+    subgraph VSCodeExtensionHost[VSCode Extension Host]
+        subgraph CoreExtension[Core Extension]
+            ExtensionEntry[Extension Entry<br/>src/extension.ts]
+            WebviewProvider[WebviewProvider<br/>src/core/webview/index.ts]
+            Controller[Controller<br/>src/core/controller/index.ts]
+            Task[Task<br/>src/core/task/index.ts]
+            GlobalState[VSCode Global State]
+            SecretsStorage[VSCode Secrets Storage]
+            McpHub[McpHub<br/>src/services/mcp/McpHub.ts]
+        end
+
+        subgraph WebviewUI[Webview UI]
+            WebviewApp[React App<br/>webview-ui/src/App.tsx]
+            ExtStateContext[ExtensionStateContext<br/>webview-ui/src/context/ExtensionStateContext.tsx]
+            ReactComponents[React Components]
+        end
+
+        subgraph Storage
+            TaskStorage[Task Storage<br/>Per-Task Files & History]
+            CheckpointSystem[Git-based Checkpoints]
+        end
+
+        subgraph apiProviders[API Providers]
+            AnthropicAPI[Anthropic]
+            OpenRouterAPI[OpenRouter]
+            BedrockAPI[AWS Bedrock]
+            OtherAPIs[Other Providers]
+        end
+
+        subgraph MCPServers[MCP Servers]
+            ExternalMcpServers[External MCP Servers]
+        end
+    end
+
+    %% Core Extension Data Flow
+    ExtensionEntry --> WebviewProvider
+    WebviewProvider --> Controller
+    Controller --> Task
+    Controller --> McpHub
+    Task --> GlobalState
+    Task --> SecretsStorage
+    Task --> TaskStorage
+    Task --> CheckpointSystem
+    Task --> |API Requests| apiProviders
+    McpHub --> |Connects to| ExternalMcpServers
+    Task --> |Uses| McpHub
+
+    %% Webview Data Flow
+    WebviewApp --> ExtStateContext
+    ExtStateContext --> ReactComponents
+
+    %% Bidirectional Communication
+    WebviewProvider <-->|postMessage| ExtStateContext
+
+    style GlobalState fill:#f9f,stroke:#333,stroke-width:2px
+    style SecretsStorage fill:#f9f,stroke:#333,stroke-width:2px
+    style ExtStateContext fill:#bbf,stroke:#333,stroke-width:2px
+    style WebviewProvider fill:#bfb,stroke:#333,stroke-width:2px
+    style McpHub fill:#bfb,stroke:#333,stroke-width:2px
+    style apiProviders fill:#fdb,stroke:#333,stroke-width:2px
+```
+
+## Definitions 
+
+- **Core Extension**: Anything inside the src folder, organized into modular components
+- **Core Extension State**: Managed by the Controller class in src/core/controller/index.ts, which serves as the single source of truth for the extension's state. It manages multiple types of persistent storage (global state, workspace state, and secrets), handles state distribution to both the core extension and webview components, and coordinates state across multiple extension instances. This includes managing API configurations, task history, settings, and MCP configurations.
+- **Webview**: Anything inside the webview-ui. All the react or view's seen by the user and user interaction components
+- **Webview State**: Managed by ExtensionStateContext in webview-ui/src/context/ExtensionStateContext.tsx, which provides React components with access to the extension's state through a context provider pattern. It maintains local state for UI components, handles real-time updates through message events, manages partial message updates, and provides methods for state modifications. The context includes extension version, messages, task history, theme, API configurations, MCP servers, marketplace catalog, and workspace file paths. It synchronizes with the core extension through VSCode's message passing system and provides type-safe access to state through a custom hook (useExtensionState).
+
+### Core Extension Architecture
+
+The core extension follows a clear hierarchical structure:
+
+1. **WebviewProvider** (src/core/webview/index.ts): Manages the webview lifecycle and communication
+2. **Controller** (src/core/controller/index.ts): Handles webview messages and task management
+3. **Task** (src/core/task/index.ts): Executes API requests and tool operations
+
+This architecture provides clear separation of concerns:
+- WebviewProvider focuses on VSCode webview integration
+- Controller manages state and coordinates tasks
+- Task handles the execution of AI requests and tool operations
+
+### WebviewProvider Implementation
+
+The WebviewProvider class in `src/core/webview/index.ts` is responsible for:
+
+- Managing multiple active instances through a static set (`activeInstances`)
+- Handling webview lifecycle events (creation, visibility changes, disposal)
+- Implementing HTML content generation with proper CSP headers
+- Supporting Hot Module Replacement (HMR) for development
+- Setting up message listeners between the webview and extension
+
+The WebviewProvider maintains a reference to the Controller and delegates message handling to it. It also handles the creation of both sidebar and tab panel webviews, allowing Cline to be used in different contexts within VSCode.
+
+### Core Extension State
+
+The `Controller` class manages multiple types of persistent storage:
+
+- **Global State:** Stored across all VSCode instances. Used for settings and data that should persist globally.
+- **Workspace State:** Specific to the current workspace. Used for task-specific data and settings.
+- **Secrets:** Secure storage for sensitive information like API keys.
+
+The `Controller` handles the distribution of state to both the core extension and webview components. It also coordinates state across multiple extension instances, ensuring consistency.
+
+State synchronization between instances is handled through:
+- File-based storage for task history and conversation data
+- VSCode's global state API for settings and configuration
+- Secrets storage for sensitive information
+- Event listeners for file changes and configuration updates
+
+The Controller implements methods for:
+- Saving and loading task state
+- Managing API configurations
+- Handling user authentication
+- Coordinating MCP server connections
+- Managing task history and checkpoints
+
+### Webview State
+
+The `ExtensionStateContext` in `webview-ui/src/context/ExtensionStateContext.tsx` provides React components with access to the extension's state. It uses a context provider pattern and maintains local state for UI components. The context includes:
+
+- Extension version
+- Messages
+- Task history
+- Theme
+- API configurations
+- MCP servers
+- Marketplace catalog
+- Workspace file paths
+
+It synchronizes with the core extension through VSCode's message passing system and provides type-safe access to the state via a custom hook (`useExtensionState`).
+
+The ExtensionStateContext handles:
+- Real-time updates through message events
+- Partial message updates for streaming content
+- State modifications through setter methods
+- Type-safe access to state through a custom hook
+
+## API Provider System
+
+Cline supports multiple AI providers through a modular API provider system. Each provider is implemented as a separate module in the `src/api/providers/` directory and follows a common interface.
+
+### API Provider Architecture
+
+The API system consists of:
+
+1. **API Handlers**: Provider-specific implementations in `src/api/providers/`
+2. **API Transformers**: Stream transformation utilities in `src/api/transform/`
+3. **API Configuration**: User settings for API keys and endpoints
+4. **API Factory**: Builder function to create the appropriate handler
+
+Key providers include:
+- **Anthropic**: Direct integration with Claude models
+- **OpenRouter**: Meta-provider supporting multiple model providers
+- **AWS Bedrock**: Integration with Amazon's AI services
+- **Gemini**: Google's AI models
+- **Ollama**: Local model hosting
+- **LM Studio**: Local model hosting
+- **VSCode LM**: VSCode's built-in language models
+
+### API Configuration Management
+
+API configurations are stored securely:
+- API keys are stored in VSCode's secrets storage
+- Model selections and non-sensitive settings are stored in global state
+- The Controller manages switching between providers and updating configurations
+
+The system supports:
+- Secure storage of API keys
+- Model selection and configuration
+- Automatic retry and error handling
+- Token usage tracking and cost calculation
+- Context window management
+
+### Plan/Act Mode API Configuration
+
+Cline supports separate model configurations for Plan and Act modes:
+- Different models can be used for planning vs. execution
+- The system preserves model selections when switching modes
+- The Controller handles the transition between modes and updates the API configuration accordingly
+
+## Task Execution System
+
+The Task class is responsible for executing AI requests and tool operations. Each task runs in its own instance of the Task class, ensuring isolation and proper state management.
+
+### Task Execution Loop
+
+The core task execution loop follows this pattern:
+
+```typescript
+class Task {
+  async initiateTaskLoop(userContent: UserContent, isNewTask: boolean) {
+    while (!this.abort) {
+      // 1. Make API request and stream response
+      const stream = this.attemptApiRequest()
+      
+      // 2. Parse and present content blocks
+      for await (const chunk of stream) {
+        switch (chunk.type) {
+          case "text":
+            // Parse into content blocks
+            this.assistantMessageContent = parseAssistantMessage(chunk.text)
+            // Present blocks to user
+            await this.presentAssistantMessage()
+            break
+        }
+      }
+      
+      // 3. Wait for tool execution to complete
+      await pWaitFor(() => this.userMessageContentReady)
+      
+      // 4. Continue loop with tool result
+      const recDidEndLoop = await this.recursivelyMakeClineRequests(
+        this.userMessageContent
+      )
+    }
+  }
+}
+```
+
+### Message Streaming System
+
+The streaming system handles real-time updates and partial content:
+
+```typescript
+class Task {
+  async presentAssistantMessage() {
+    // Handle streaming locks to prevent race conditions
+    if (this.presentAssistantMessageLocked) {
+      this.presentAssistantMessageHasPendingUpdates = true
+      return
+    }
+    this.presentAssistantMessageLocked = true
+
+    // Present current content block
+    const block = this.assistantMessageContent[this.currentStreamingContentIndex]
+    
+    // Handle different types of content
+    switch (block.type) {
+      case "text":
+        await this.say("text", content, undefined, block.partial)
+        break
+      case "tool_use":
+        // Handle tool execution
+        break
+    }
+
+    // Move to next block if complete
+    if (!block.partial) {
+      this.currentStreamingContentIndex++
+    }
+  }
+}
+```
+
+### Tool Execution Flow
+
+Tools follow a strict execution pattern:
+
+```typescript
+class Task {
+  async executeToolWithApproval(block: ToolBlock) {
+    // 1. Check auto-approval settings
+    if (this.shouldAutoApproveTool(block.name)) {
+      await this.say("tool", message)
+      this.consecutiveAutoApprovedRequestsCount++
+    } else {
+      // 2. Request user approval
+      const didApprove = await askApproval("tool", message)
+      if (!didApprove) {
+        this.didRejectTool = true
+        return
+      }
+    }
+
+    // 3. Execute tool
+    const result = await this.executeTool(block)
+
+    // 4. Save checkpoint
+    await this.saveCheckpoint()
+
+    // 5. Return result to API
+    return result
+  }
+}
+```
+
+### Error Handling & Recovery
+
+The system includes robust error handling:
+
+```typescript
+class Task {
+  async handleError(action: string, error: Error) {
+    // 1. Check if task was abandoned
+    if (this.abandoned) return
+    
+    // 2. Format error message
+    const errorString = `Error ${action}: ${error.message}`
+    
+    // 3. Present error to user
+    await this.say("error", errorString)
+    
+    // 4. Add error to tool results
+    pushToolResult(formatResponse.toolError(errorString))
+    
+    // 5. Cleanup resources
+    await this.diffViewProvider.revertChanges()
+    await this.browserSession.closeBrowser()
+  }
+}
+```
+
+### API Request & Token Management
+
+The Task class handles API requests with built-in retry, streaming, and token management:
+
+```typescript
+class Task {
+  async *attemptApiRequest(previousApiReqIndex: number): ApiStream {
+    // 1. Wait for MCP servers to connect
+    await pWaitFor(() => this.controllerRef.deref()?.mcpHub?.isConnecting !== true)
+
+    // 2. Manage context window
+    const previousRequest = this.clineMessages[previousApiReqIndex]
+    if (previousRequest?.text) {
+      const { tokensIn, tokensOut } = JSON.parse(previousRequest.text || "{}")
+      const totalTokens = (tokensIn || 0) + (tokensOut || 0)
+      
+      // Truncate conversation if approaching context limit
+      if (totalTokens >= maxAllowedSize) {
+        this.conversationHistoryDeletedRange = this.contextManager.getNextTruncationRange(
+          this.apiConversationHistory,
+          this.conversationHistoryDeletedRange,
+          totalTokens / 2 > maxAllowedSize ? "quarter" : "half"
+        )
+      }
+    }
+
+    // 3. Handle streaming with automatic retry
+    try {
+      this.isWaitingForFirstChunk = true
+      const firstChunk = await iterator.next()
+      yield firstChunk.value
+      this.isWaitingForFirstChunk = false
+      
+      // Stream remaining chunks
+      yield* iterator
+    } catch (error) {
+      // 4. Error handling with retry
+      if (isOpenRouter && !this.didAutomaticallyRetryFailedApiRequest) {
+        await setTimeoutPromise(1000)
+        this.didAutomaticallyRetryFailedApiRequest = true
+        yield* this.attemptApiRequest(previousApiReqIndex)
+        return
+      }
+      
+      // 5. Ask user to retry if automatic retry failed
+      const { response } = await this.ask(
+        "api_req_failed",
+        this.formatErrorWithStatusCode(error)
+      )
+      if (response === "yesButtonClicked") {
+        await this.say("api_req_retried")
+        yield* this.attemptApiRequest(previousApiReqIndex)
+        return
+      }
+    }
+  }
+}
+```
+
+Key features:
+
+1. **Context Window Management**
+   - Tracks token usage across requests
+   - Automatically truncates conversation when needed
+   - Preserves important context while freeing space
+   - Handles different model context sizes
+
+2. **Streaming Architecture**
+   - Real-time chunk processing
+   - Partial content handling
+   - Race condition prevention
+   - Error recovery during streaming
+
+3. **Error Handling**
+   - Automatic retry for transient failures
+   - User-prompted retry for persistent issues
+   - Detailed error reporting
+   - State cleanup on failure
+
+4. **Token Tracking**
+   - Per-request token counting
+   - Cumulative usage tracking
+   - Cost calculation
+   - Cache hit monitoring
+
+### Context Management System
+
+The Context Management System handles conversation history truncation to prevent context window overflow errors. Implemented in the `ContextManager` class, it ensures long-running conversations remain within model context limits while preserving critical context.
+
+Key features:
+
+1. **Model-Aware Sizing**: Dynamically adjusts based on different model context windows (64K for DeepSeek, 128K for most models, 200K for Claude).
+
+2. **Proactive Truncation**: Monitors token usage and preemptively truncates conversations when approaching limits, maintaining buffers of 27K-40K tokens depending on the model.
+
+3. **Intelligent Preservation**: Always preserves the original task message and maintains the user-assistant conversation structure when truncating.
+
+4. **Adaptive Strategies**: Uses different truncation strategies based on context pressure - removing half of the conversation for moderate pressure or three-quarters for severe pressure.
+
+5. **Error Recovery**: Includes specialized detection for context window errors from different providers with automatic retry and more aggressive truncation when needed.
+
+### Task State & Resumption
+
+The Task class provides robust task state management and resumption capabilities:
+
+```typescript
+class Task {
+  async resumeTaskFromHistory() {
+    // 1. Load saved state
+    this.clineMessages = await getSavedClineMessages(this.getContext(), this.taskId)
+    this.apiConversationHistory = await getSavedApiConversationHistory(this.getContext(), this.taskId)
+
+    // 2. Handle interrupted tool executions
+    const lastMessage = this.apiConversationHistory[this.apiConversationHistory.length - 1]
+    if (lastMessage.role === "assistant") {
+      const toolUseBlocks = content.filter(block => block.type === "tool_use")
+      if (toolUseBlocks.length > 0) {
+        // Add interrupted tool responses
+        const toolResponses = toolUseBlocks.map(block => ({
+          type: "tool_result",
+          tool_use_id: block.id,
+          content: "Task was interrupted before this tool call could be completed."
+        }))
+        modifiedOldUserContent = [...toolResponses]
+      }
+    }
+
+    // 3. Notify about interruption
+    const agoText = this.getTimeAgoText(lastMessage?.ts)
+    newUserContent.push({
+      type: "text",
+      text: `[TASK RESUMPTION] This task was interrupted ${agoText}. It may or may not be complete, so please reassess the task context.`
+    })
+
+    // 4. Resume task execution
+    await this.initiateTaskLoop(newUserContent, false)
+  }
+
+  private async saveTaskState() {
+    // Save conversation history
+    await saveApiConversationHistory(this.getContext(), this.taskId, this.apiConversationHistory)
+    await saveClineMessages(this.getContext(), this.taskId, this.clineMessages)
+    
+    // Create checkpoint
+    const commitHash = await this.checkpointTracker?.commit()
+    
+    // Update task history
+    await this.controllerRef.deref()?.updateTaskHistory({
+      id: this.taskId,
+      ts: lastMessage.ts,
+      task: taskMessage.text,
+      // ... other metadata
+    })
+  }
+}
+```
+
+Key aspects of task state management:
+
+1. **Task Persistence**
+   - Each task has a unique ID and dedicated storage directory
+   - Conversation history is saved after each message
+   - File changes are tracked through Git-based checkpoints
+   - Terminal output and browser state are preserved
+
+2. **State Recovery**
+   - Tasks can be resumed from any point
+   - Interrupted tool executions are handled gracefully
+   - File changes can be restored from checkpoints
+   - Context is preserved across VSCode sessions
+
+3. **Workspace Synchronization**
+   - File changes are tracked through Git
+   - Checkpoints are created after tool executions
+   - State can be restored to any checkpoint
+   - Changes can be compared between checkpoints
+
+4. **Error Recovery**
+   - Failed API requests can be retried
+   - Interrupted tool executions are marked
+   - Resources are cleaned up properly
+   - User is notified of state changes
+
+## Plan/Act Mode System
+
+Cline implements a dual-mode system that separates planning from execution:
+
+### Mode Architecture
+
+The Plan/Act mode system consists of:
+
+1. **Mode State**: Stored in `chatSettings.mode` in the Controller's state
+2. **Mode Switching**: Handled by `togglePlanActModeWithChatSettings` in the Controller
+3. **Mode-specific Models**: Optional configuration to use different models for each mode
+4. **Mode-specific Prompting**: Different system prompts for planning vs. execution
+
+### Mode Switching Process
+
+When switching between modes:
+
+1. The current model configuration is saved to mode-specific state
+2. The previous mode's model configuration is restored
+3. The Task instance is updated with the new mode
+4. The webview is notified of the mode change
+5. Telemetry events are captured for analytics
+
+### Plan Mode
+
+Plan mode is designed for:
+- Information gathering and context building
+- Asking clarifying questions
+- Creating detailed execution plans
+- Discussing approaches with the user
+
+In Plan mode, the AI uses the `plan_mode_respond` tool to engage in conversational planning without executing actions.
+
+### Act Mode
+
+Act mode is designed for:
+- Executing the planned actions
+- Using tools to modify files, run commands, etc.
+- Implementing the solution
+- Providing results and completion feedback
+
+In Act mode, the AI has access to all tools except `plan_mode_respond` and focuses on implementation rather than discussion.
+
+## Data Flow & State Management
+
+### Core Extension Role
+
+The Controller acts as the single source of truth for all persistent state. It:
+- Manages VSCode global state and secrets storage
+- Coordinates state updates between components
+- Ensures state consistency across webview reloads
+- Handles task-specific state persistence
+- Manages checkpoint creation and restoration
+
+### Terminal Management
+
+The Task class manages terminal instances and command execution:
+
+```typescript
+class Task {
+  async executeCommandTool(command: string): Promise<[boolean, ToolResponse]> {
+    // 1. Get or create terminal
+    const terminalInfo = await this.terminalManager.getOrCreateTerminal(cwd)
+    terminalInfo.terminal.show()
+
+    // 2. Execute command with output streaming
+    const process = this.terminalManager.runCommand(terminalInfo, command)
+    
+    // 3. Handle real-time output
+    let result = ""
+    process.on("line", (line) => {
+      result += line + "\n"
+      if (!didContinue) {
+        sendCommandOutput(line)
+      } else {
+        this.say("command_output", line)
+      }
+    })
+
+    // 4. Wait for completion or user feedback
+    let completed = false
+    process.once("completed", () => {
+      completed = true
+    })
+
+    await process
+
+    // 5. Return result
+    if (completed) {
+      return [false, `Command executed.\n${result}`]
+    } else {
+      return [
+        false,
+        `Command is still running in the user's terminal.\n${result}\n\nYou will be updated on the terminal status and new output in the future.`
+      ]
+    }
+  }
+}
+```
+
+Key features:
+1. **Terminal Instance Management**
+   - Multiple terminal support
+   - Terminal state tracking (busy/inactive)
+   - Process cooldown monitoring
+   - Output history per terminal
+
+2. **Command Execution**
+   - Real-time output streaming
+   - User feedback handling
+   - Process state monitoring
+   - Error recovery
+
+### Browser Session Management
+
+The Task class handles browser automation through Puppeteer:
+
+```typescript
+class Task {
+  async executeBrowserAction(action: BrowserAction): Promise<BrowserActionResult> {
+    switch (action) {
+      case "launch":
+        // 1. Launch browser with fixed resolution
+        await this.browserSession.launchBrowser()
+        return await this.browserSession.navigateToUrl(url)
+
+      case "click":
+        // 2. Handle click actions with coordinates
+        return await this.browserSession.click(coordinate)
+
+      case "type":
+        // 3. Handle keyboard input
+        return await this.browserSession.type(text)
+
+      case "close":
+        // 4. Clean up resources
+        return await this.browserSession.closeBrowser()
+    }
+  }
+}
+```
+
+Key aspects:
+1. **Browser Control**
+   - Fixed 900x600 resolution window
+   - Single instance per task lifecycle
+   - Automatic cleanup on task completion
+   - Console log capture
+
+2. **Interaction Handling**
+   - Coordinate-based clicking
+   - Keyboard input simulation
+   - Screenshot capture
+   - Error recovery
+
+## MCP (Model Context Protocol) Integration
+
+### MCP Architecture
+
+The MCP system consists of:
+
+1. **McpHub Class**: Central manager in `src/services/mcp/McpHub.ts`
+2. **MCP Connections**: Manages connections to external MCP servers
+3. **MCP Settings**: Configuration stored in a JSON file
+4. **MCP Marketplace**: Online catalog of available MCP servers
+5. **MCP Tools & Resources**: Capabilities exposed by connected servers
+
+The McpHub class:
+- Manages the lifecycle of MCP server connections
+- Handles server configuration through a settings file
+- Provides methods for calling tools and accessing resources
+- Implements auto-approval settings for MCP tools
+- Monitors server health and handles reconnection
+
+### MCP Server Types
+
+Cline supports two types of MCP server connections:
+- **Stdio**: Command-line based servers that communicate via standard I/O
+- **SSE**: HTTP-based servers that communicate via Server-Sent Events
+
+### MCP Server Management
+
+The McpHub class provides methods for:
+- Discovering and connecting to MCP servers
+- Monitoring server health and status
+- Restarting servers when needed
+- Managing server configurations
+- Setting timeouts and auto-approval rules
+
+### MCP Tool Integration
+
+MCP tools are integrated into the Task execution system:
+- Tools are discovered and registered at connection time
+- The Task class can call MCP tools through the McpHub
+- Tool results are streamed back to the AI
+- Auto-approval settings can be configured per tool
+
+### MCP Marketplace
+
+The MCP Marketplace provides:
+- A catalog of available MCP servers
+- One-click installation
+- README previews
+- Server status monitoring
+
+The Controller class manages MCP servers through the McpHub service:
+
+```typescript
+class Controller {
+  mcpHub?: McpHub
+
+  constructor(context: vscode.ExtensionContext, outputChannel: vscode.OutputChannel, webviewProvider: WebviewProvider) {
+    this.mcpHub = new McpHub(this)
+  }
+
+  async downloadMcp(mcpId: string) {
+    // Fetch server details from marketplace
+    const response = await axios.post<McpDownloadResponse>(
+      "https://api.cline.bot/v1/mcp/download",
+      { mcpId },
+      {
+        headers: { "Content-Type": "application/json" },
+        timeout: 10000,
+      }
+    )
+
+    // Create task with context from README
+    const task = `Set up the MCP server from ${mcpDetails.githubUrl}...`
+
+    // Initialize task and show chat view
+    await this.initClineWithTask(task)
+  }
+}
+```
+
+## Conclusion
+
+This guide provides a comprehensive overview of the Cline extension architecture, with special focus on state management, data persistence, and code organization. Following these patterns ensures robust feature implementation with proper state handling across the extension's components.
+
+Remember:
+- Always persist important state in the extension
+- The core extension follows a WebviewProvider -> Controller -> Task flow
+- Use proper typing for all state and messages
+- Handle errors and edge cases
+- Test state persistence across webview reloads
+- Follow the established patterns for consistency
+- Place new code in appropriate directories
+- Maintain clear separation of concerns
+- Install dependencies in correct package.json
+
+## Contributing
+
+Contributions to the Cline extension are welcome! Please follow these guidelines:
+
+When adding new tools or API providers, follow the existing patterns in the `src/integrations/` and `src/api/providers/` directories, respectively. Ensure that your code is well-documented and includes appropriate error handling.
+
+The `.clineignore` file allows users to specify files and directories that Cline should not access. When implementing new features, respect the `.clineignore` rules and ensure that your code does not attempt to read or modify ignored files.

.codespellrc

@@ -0,0 +1,6 @@
+[codespell]
+# Ref: https://github.com/codespell-project/codespell#using-a-config-file
+skip = .git*,*.svg,package-lock.json,*.css,.codespellrc,locales
+check-hidden = true
+ignore-regex = (\b(optIn|isTaller)\b|https://\S+)
+# ignore-words-list =

.github/CODEOWNERS

@@ -1 +1 @@
-*       @saoudrizwan @ocasta181 @NightTrek @pashpashpash
+*       @saoudrizwan @ocasta181 @NightTrek @pashpashpash @dcbartlett @saito-sv @Garoth

.github/ISSUE_TEMPLATE/bug_report.yml

@@ -31,6 +31,14 @@ body:
           label: Relevant API REQUEST output
           description: Please copy and paste any relevant output. This will be automatically formatted into code, so no need for backticks.
           render: shell
+    - type: input
+      id: provider-model
+      attributes:
+          label: Provider/Model
+          description: What provider and model were you using when the issue occurred?
+          placeholder: "e.g., cline:anthropic/claude-3.7-sonnet, gemini:gemini-2.5-pro-exp-03-25"
+      validations:
+          required: true
     - type: input
       id: operating-system
       attributes:

.github/dependabot.yml

@@ -28,10 +28,7 @@ updates:
               patterns:
                   - "*"
       ignore:
-          # Ignore CRA and related packages that often have false positives
-          - dependency-name: "react-scripts"
           - dependency-name: "@testing-library/*"
-          - dependency-name: "web-vitals"
           - dependency-name: "*"
             update-types:
                 - "version-update:semver-major"

.github/pull_request_template.md

@@ -13,7 +13,10 @@
 -   [ ] 🐛 Bug fix (non-breaking change which fixes an issue)
 -   [ ] ✨ New feature (non-breaking change which adds functionality)
 -   [ ] 💥 Breaking change (fix or feature that would cause existing functionality to not work as expected)
+-   [ ] ♻️ Refactor Changes
+-   [ ] 💅 Cosmetic Changes
 -   [ ] 📚 Documentation update
+-   [ ] 🏃 Workflow Changes
 
 ### Pre-flight Checklist
 
@@ -21,6 +24,7 @@
 
 -   [ ] Changes are limited to a single feature, bugfix or chore (split larger changes into separate PRs)
 -   [ ] Tests are passing (`npm test`) and code is formatted and linted (`npm run format && npm run lint`)
+-   [ ] I have created a changeset using `npm run changeset` (required for user-facing changes)
 -   [ ] I have reviewed [contributor guidelines](https://github.com/cline/cline/blob/main/CONTRIBUTING.md)
 
 ### Screenshots

.github/scripts/coverage_check/__init__.py

@@ -0,0 +1,19 @@
+"""
+Coverage utility package for GitHub Actions workflows.
+This package handles extracting coverage percentages, comparing them, and generating PR comments.
+"""
+
+# Import external dependencies
+import requests
+
+# Import main function for CLI usage
+from .__main__ import main
+
+# Import functions from extraction module
+from .extraction import extract_coverage, compare_coverage, run_coverage, set_verbose
+
+# Import functions from github_api module
+from .github_api import generate_comment, post_comment, set_github_output
+
+# Import functions from workflow module
+from .workflow import process_coverage_workflow

.github/scripts/coverage_check/__main__.py

@@ -0,0 +1,154 @@
+"""
+Main module.
+This module provides the CLI interface for the coverage utility script.
+"""
+
+import sys
+import argparse
+
+from .extraction import extract_coverage, compare_coverage, run_coverage, set_verbose
+from .github_api import generate_comment, post_comment, set_github_output
+from .workflow import process_coverage_workflow
+from .util import log
+
+def setup_verbose_mode(args):
+    """
+    Set up verbose mode based on command line arguments.
+    
+    Args:
+        args: Parsed command line arguments
+    """
+    if getattr(args, 'verbose', False):
+        set_verbose(True)
+        log("Verbose mode enabled")
+
+def main():
+    # Create parent parser with common arguments
+    parent_parser = argparse.ArgumentParser(add_help=False)
+    parent_parser.add_argument('-v', '--verbose', action='store_true', help='Enable verbose output')
+
+    # Create main parser that inherits common arguments
+    parser = argparse.ArgumentParser(description='Coverage utility script for GitHub Actions workflows', parents=[parent_parser])
+    subparsers = parser.add_subparsers(dest='command', help='Command to run')
+    
+    # extract-coverage command - used directly in workflow
+    extract_parser = subparsers.add_parser('extract-coverage', help='Extract coverage percentage from a file', parents=[parent_parser])
+    extract_parser.add_argument('file_path', help='Path to the coverage report file')
+    extract_parser.add_argument('--type', choices=['extension', 'webview'], default='extension',
+                               help='Type of coverage report')
+    extract_parser.add_argument('--github-output', action='store_true', help='Output in GitHub Actions format')
+    
+    # compare-coverage command - used by process-workflow
+    compare_parser = subparsers.add_parser('compare-coverage', help='Compare coverage percentages', parents=[parent_parser])
+    compare_parser.add_argument('base_cov', help='Base branch coverage percentage')
+    compare_parser.add_argument('pr_cov', help='PR branch coverage percentage')
+    compare_parser.add_argument('--output-prefix', default='', help='Prefix for GitHub Actions output variables')
+    compare_parser.add_argument('--github-output', action='store_true', help='Output in GitHub Actions format')
+    
+    # generate-comment command - used by process-workflow
+    comment_parser = subparsers.add_parser('generate-comment', help='Generate PR comment with coverage comparison', parents=[parent_parser])
+    comment_parser.add_argument('base_ext_cov', help='Base branch extension coverage')
+    comment_parser.add_argument('pr_ext_cov', help='PR branch extension coverage')
+    comment_parser.add_argument('ext_decreased', help='Whether extension coverage decreased (true/false)')
+    comment_parser.add_argument('ext_diff', help='Extension coverage difference')
+    comment_parser.add_argument('base_web_cov', help='Base branch webview coverage')
+    comment_parser.add_argument('pr_web_cov', help='PR branch webview coverage')
+    comment_parser.add_argument('web_decreased', help='Whether webview coverage decreased (true/false)')
+    comment_parser.add_argument('web_diff', help='Webview coverage difference')
+    
+    # post-comment command - used by process-workflow
+    post_parser = subparsers.add_parser('post-comment', help='Post a comment to a GitHub PR', parents=[parent_parser])
+    post_parser.add_argument('comment_path', help='Path to the file containing the comment text')
+    post_parser.add_argument('pr_number', help='PR number')
+    post_parser.add_argument('repo', help='Repository in the format "owner/repo"')
+    post_parser.add_argument('--token', help='GitHub token')
+    
+    # run-coverage command - used by process-workflow
+    run_parser = subparsers.add_parser('run-coverage', help='Run a coverage command and extract the coverage percentage', parents=[parent_parser])
+    run_parser.add_argument('coverage_cmd', help='Command to run')
+    run_parser.add_argument('output_file', help='File to save the output to')
+    run_parser.add_argument('--type', choices=['extension', 'webview'], default='extension',
+                           help='Type of coverage report')
+    run_parser.add_argument('--github-output', action='store_true', help='Output in GitHub Actions format')
+    
+    # process-workflow command - used directly in workflow
+    workflow_parser = subparsers.add_parser('process-workflow', help='Process the entire coverage workflow', parents=[parent_parser])
+    workflow_parser.add_argument('--base-branch', required=True, help='Base branch name')
+    workflow_parser.add_argument('--pr-number', help='PR number')
+    workflow_parser.add_argument('--repo', help='Repository in the format "owner/repo"')
+    workflow_parser.add_argument('--token', help='GitHub token')
+    
+    # set-github-output command - used by process-workflow
+    output_parser = subparsers.add_parser('set-github-output', help='Set GitHub Actions output variable', parents=[parent_parser])
+    output_parser.add_argument('name', help='Output variable name')
+    output_parser.add_argument('value', help='Output variable value')
+    
+    args = parser.parse_args()
+    
+    # Set up verbose mode
+    setup_verbose_mode(args)
+    
+    if args.command == 'extract-coverage':
+        log(f"Extracting coverage from file: {args.file_path} (type: {args.type})")
+        coverage_pct = extract_coverage(args.file_path, args.type)
+        if args.github_output:
+            set_github_output(f"{args.type}_coverage", coverage_pct)
+        else:
+            log(f"Coverage: {coverage_pct}%")
+        
+    elif args.command == 'compare-coverage':
+        log(f"Comparing coverage: base={args.base_cov}%, PR={args.pr_cov}%")
+        decreased, diff = compare_coverage(args.base_cov, args.pr_cov)
+        if args.github_output:
+            prefix = args.output_prefix
+            set_github_output(f"{prefix}decreased", str(decreased).lower())
+            set_github_output(f"{prefix}diff", diff)
+            log(f"Coverage difference: {diff}%")
+            log(f"Coverage decreased: {decreased}")
+        else:
+            log(f"decreased={str(decreased).lower()}")
+            log(f"diff={diff}")
+        
+    elif args.command == 'generate-comment':
+        log("Generating coverage comparison comment")
+        comment = generate_comment(
+            args.base_ext_cov, args.pr_ext_cov, args.ext_decreased, args.ext_diff,
+            args.base_web_cov, args.pr_web_cov, args.web_decreased, args.web_diff
+        )
+        # Output the comment to stdout
+        log(comment)
+        
+    elif args.command == 'post-comment':
+        log(f"Posting comment from {args.comment_path} to PR #{args.pr_number} in {args.repo}")
+        post_comment(args.comment_path, args.pr_number, args.repo, args.token)
+        
+    elif args.command == 'run-coverage':
+        log(f"Running coverage command: {args.coverage_cmd}")
+        log(f"Output file: {args.output_file}")
+        log(f"Coverage type: {args.type}")
+        coverage_pct = run_coverage(args.coverage_cmd, args.output_file, args.type)
+        if args.github_output:
+            set_github_output(f"{args.type}_coverage", coverage_pct)
+        else:
+            log(f"Coverage: {coverage_pct}%")
+        
+    elif args.command == 'process-workflow':
+        log("Processing coverage workflow")
+        log(f"Base branch: {args.base_branch}")
+        if args.pr_number:
+            log(f"PR number: {args.pr_number}")
+        if args.repo:
+            log(f"Repository: {args.repo}")
+        process_coverage_workflow(args)
+        
+    elif args.command == 'set-github-output':
+        log(f"Setting GitHub output: {args.name}={args.value}")
+        set_github_output(args.name, args.value)
+    
+    else:
+        log("No command specified")
+        parser.print_help()
+        sys.exit(1)
+
+if __name__ == "__main__":
+    main()

.github/scripts/coverage_check/extraction.py

@@ -0,0 +1,265 @@
+"""
+Coverage extraction module.
+This module handles extracting coverage percentages from coverage report files.
+"""
+
+import os
+import re
+import sys
+import shlex
+import subprocess
+import traceback
+from .util import log, file_exists, get_file_size, list_directory, is_safe_command, run_command
+
+# Global verbose flag
+verbose = False
+
+def set_verbose(value):
+    """Set the global verbose flag."""
+    global verbose
+    verbose = value
+
+def print_debug_output(content, coverage_type):
+    """
+    Print debug information about the coverage output.
+    
+    Args:
+        content: The content of the coverage file
+        coverage_type: Type of coverage report (extension or webview)
+    """
+    if not verbose:
+        return
+
+    # Extract and print only the coverage summary section
+    if coverage_type == "extension":
+        # Look for the coverage summary section
+        summary_match = re.search(r'=============================== Coverage summary ===============================\n(.*?)\n=+', content, re.DOTALL)
+        if summary_match:
+            sys.stdout.write("\n##[group]EXTENSION COVERAGE SUMMARY\n")
+            sys.stdout.write("=============================== Coverage summary ===============================\n")
+            sys.stdout.write(summary_match.group(1) + "\n")
+            sys.stdout.write("================================================================================\n")
+            sys.stdout.write("##[endgroup]\n")
+            sys.stdout.flush()
+        else:
+            sys.stdout.write("\n##[warning]No coverage summary found in extension coverage file\n")
+            sys.stdout.flush()
+    else:  # webview
+        # Look for the coverage table - specifically the "All files" row
+        table_match = re.search(r'% Coverage report from v8.*?-+\|.*?\n.*?\n(All files.*?)(?:\n[^\n]*\|)', content, re.DOTALL)
+        if table_match:
+            sys.stdout.write("\n##[group]WEBVIEW COVERAGE SUMMARY\n")
+            sys.stdout.write("% Coverage report from v8\n")
+            sys.stdout.write("-------------------|---------|----------|---------|---------|-------------------\n")
+            sys.stdout.write("File               | % Stmts | % Branch | % Funcs | % Lines | Uncovered Line #s \n")
+            sys.stdout.write("-------------------|---------|----------|---------|---------|-------------------\n")
+            sys.stdout.write(table_match.group(1) + "\n")
+            sys.stdout.write("-------------------|---------|----------|---------|---------|-------------------\n")
+            sys.stdout.write("##[endgroup]\n")
+            sys.stdout.flush()
+        else:
+            sys.stdout.write("\n##[warning]No coverage table found in webview coverage file\n")
+            sys.stdout.flush()
+
+def extract_coverage(file_path, coverage_type="extension"):
+    """
+    Extract coverage percentage from a coverage report file.
+    
+    Args:
+        file_path: Path to the coverage report file
+        coverage_type: Type of coverage report (extension or webview)
+        
+    Returns:
+        Coverage percentage as a float
+    """
+    
+    # Always print file path for debugging
+    log(f"Checking coverage file: {file_path}")
+    
+    # Check if file exists and get its size
+    if not file_exists(file_path):
+        sys.stdout.write(f"\n##[error]File {file_path} does not exist\n")
+        sys.stdout.flush()
+        log(f"Error: File {file_path} does not exist")
+        
+        # Check if the directory exists
+        dir_path = os.path.dirname(file_path)
+        if not os.path.exists(dir_path):
+            sys.stdout.write(f"\n##[error]Directory {dir_path} does not exist\n")
+            sys.stdout.flush()
+            log(f"Error: Directory {dir_path} does not exist")
+        else:
+            # List directory contents for debugging
+            log(f"Directory {dir_path} exists, listing contents:")
+            try:
+                dir_contents = list_directory(dir_path)
+                for name, size in dir_contents:
+                    log(f"  {name} - {size}")
+                    sys.stdout.write(f"  {name} - {size}\n")
+                sys.stdout.flush()
+            except Exception as e:
+                log(f"Error listing directory: {e}")
+        
+        return 0.0
+    
+    file_size = get_file_size(file_path)
+    log(f"File size: {file_size} bytes")
+    sys.stdout.write(f"\n##[info]Coverage file {file_path} exists, size: {file_size} bytes\n")
+    sys.stdout.flush()
+    
+    if file_size == 0:
+        sys.stdout.write(f"\n##[warning]File {file_path} is empty\n")
+        sys.stdout.flush()
+        log(f"Warning: File {file_path} is empty")
+        return 0.0
+    
+    # List directory contents for debugging
+    dir_path = os.path.dirname(file_path)
+    log(f"Directory contents of {dir_path}:")
+    try:
+        dir_contents = list_directory(dir_path)
+        for name, size in dir_contents:
+            log(f"  {name} - {size}")
+    except Exception as e:
+        log(f"Error listing directory: {e}")
+    
+    with open(file_path, 'r') as f:
+        content = f.read()
+    
+    # Print debug information if verbose
+    print_debug_output(content, coverage_type)
+    
+    # Extract coverage percentage based on coverage type
+    if coverage_type == "extension":
+        # Extract the percentage from the "Lines" row in the coverage summary
+        # Pattern: Lines : xx.xx% ( xxxxxxx/xxxxxxx )
+        lines_match = re.search(r'Lines\s*:\s*(\d+\.\d+)%', content)
+        if lines_match:
+            coverage_pct = float(lines_match.group(1))
+            if verbose:
+                sys.stdout.write(f"Pattern matched (Lines percentage): {coverage_pct}\n")
+                sys.stdout.flush()
+            return coverage_pct
+        else:
+            # No coverage data found, log full content for debugging
+            log("No coverage data found. Full file content:")
+            log("=== Full file content ===")
+            log(content)
+            log("=== End file content ===")
+    else:  # webview
+        # Extract the percentage from the "% Lines" column in the "All files" row
+        # Pattern: All files | xx.xx | xx.xx | xx.xx | xx.xx |
+        all_files_match = re.search(r'All files\s+\|\s+\d+\.\d+\s+\|\s+\d+\.\d+\s+\|\s+\d+\.\d+\s+\|\s+(\d+\.\d+)', content)
+        if all_files_match:
+            coverage_pct = float(all_files_match.group(1))
+            if verbose:
+                sys.stdout.write(f"Pattern matched (All files % Lines): {coverage_pct}\n")
+                sys.stdout.flush()
+            return coverage_pct
+        else:
+            # No coverage data found, log full content for debugging
+            log("No coverage data found. Full file content:")
+            log("=== Full file content ===")
+            log(content)
+            log("=== End file content ===")
+    
+    # If no match found, return 0.0
+    return 0.0
+
+def compare_coverage(base_cov, pr_cov):
+    """
+    Compare coverage percentages between base and PR branches.
+    
+    Args:
+        base_cov: Base branch coverage percentage
+        pr_cov: PR branch coverage percentage
+        
+    Returns:
+        Tuple of (decreased, diff)
+    """
+    try:
+        base_cov = float(base_cov)
+        pr_cov = float(pr_cov)
+    except ValueError:
+        sys.stdout.write(f"Error: Invalid coverage values - base: {base_cov}, PR: {pr_cov}\n")
+        sys.stdout.flush()
+        return False, 0
+    
+    diff = pr_cov - base_cov
+    decreased = diff < 0
+    
+    return decreased, abs(diff)
+
+def run_coverage(command, output_file, coverage_type="extension"):
+    """
+    Run a coverage command and extract the coverage percentage.
+    
+    Args:
+        command: Command to run
+        output_file: File to save the output to
+        coverage_type: Type of coverage report (extension or webview)
+        
+    Returns:
+        Coverage percentage as a float
+    
+    Raises:
+        SystemExit: If the output file is not created or is empty
+    """
+    
+    try:
+        # Run the command and capture output
+        if not is_safe_command(command):
+            error_msg = f"ERROR: Unsafe command detected: {command}"
+            log(error_msg)
+            sys.stdout.write(f"\n##[error]{error_msg}\n")
+            sys.stdout.flush()
+            sys.exit(1)
+
+        # Run command using safe execution from util
+        returncode, stdout, stderr = run_command(command)
+        
+        # Log command result
+        log(f"Command exit code: {returncode}")
+        log(f"Command stdout length: {len(stdout)} bytes")
+        log(f"Command stderr length: {len(stderr)} bytes")
+        
+        # Save output to file
+        log(f"Saving command output to {output_file}")
+        with open(output_file, 'w') as f:
+            f.write(stdout)
+            if stderr:
+                f.write("\n\n=== STDERR ===\n")
+                f.write(stderr)
+        
+        # Verify file was created and has content
+        if not file_exists(output_file):
+            error_msg = f"ERROR: Output file {output_file} was not created"
+            log(error_msg)
+            sys.stdout.write(f"\n##[error]{error_msg}\n")
+            sys.stdout.flush()
+            sys.exit(1)  # Exit with error code to fail the workflow
+            
+        file_size = get_file_size(output_file)
+        if file_size == 0:
+            error_msg = f"ERROR: Output file {output_file} is empty"
+            log(error_msg)
+            sys.stdout.write(f"\n##[error]{error_msg}\n")
+            sys.stdout.flush()
+            sys.exit(1)  # Exit with error code to fail the workflow
+            
+        log(f"Output file size: {file_size} bytes")
+        
+        # Extract coverage percentage
+        coverage_pct = extract_coverage(output_file, coverage_type)
+        
+        log(f"{coverage_type.capitalize()} coverage: {coverage_pct}%")
+        return coverage_pct
+    
+    except Exception as e:
+        error_msg = f"Error running coverage command: {e}"
+        log(error_msg)
+        sys.stdout.write(f"\n##[error]{error_msg}\n")
+        sys.stdout.flush()
+        # Print stack trace for debugging
+        log(traceback.format_exc())
+        sys.exit(1)  # Exit with error code to fail the workflow

.github/scripts/coverage_check/github_api.py

@@ -0,0 +1,177 @@
+"""
+GitHub API module.
+This module handles interactions with the GitHub API for posting comments to PRs.
+"""
+
+import os
+import requests
+from .util import log, file_exists
+
+def generate_comment(base_ext_cov, pr_ext_cov, ext_decreased, ext_diff, 
+                    base_web_cov, pr_web_cov, web_decreased, web_diff):
+    """
+    Generate a PR comment with coverage comparison.
+    
+    Args:
+        base_ext_cov: Base branch extension coverage
+        pr_ext_cov: PR branch extension coverage
+        ext_decreased: Whether extension coverage decreased
+        ext_diff: Extension coverage difference
+        base_web_cov: Base branch webview coverage
+        pr_web_cov: PR branch webview coverage
+        web_decreased: Whether webview coverage decreased
+        web_diff: Webview coverage difference
+        
+    Returns:
+        Comment text
+    """
+    from datetime import datetime
+    
+    # Convert string inputs to appropriate types
+    try:
+        base_ext_cov = float(base_ext_cov)
+        pr_ext_cov = float(pr_ext_cov)
+        # Handle ext_decreased as either string or boolean
+        if isinstance(ext_decreased, str):
+            ext_decreased = ext_decreased.lower() == 'true'
+        else:
+            ext_decreased = bool(ext_decreased)
+        ext_diff = float(ext_diff)
+        base_web_cov = float(base_web_cov)
+        pr_web_cov = float(pr_web_cov)
+        # Handle web_decreased as either string or boolean
+        if isinstance(web_decreased, str):
+            web_decreased = web_decreased.lower() == 'true'
+        else:
+            web_decreased = bool(web_decreased)
+        web_diff = float(web_diff)
+    except ValueError as e:
+        log(f"Error converting input values: {e}")
+        return ""
+    
+    # Add a unique identifier to find this comment later
+    comment = '<!-- COVERAGE_REPORT -->\n'
+    comment += '## Coverage Report\n\n'
+
+    # Extension coverage
+    comment += '### Extension Coverage\n\n'
+    comment += f'Base branch: {base_ext_cov:.0f}%\n\n'
+    comment += f'PR branch: {pr_ext_cov:.0f}%\n\n'
+
+    if ext_decreased:
+        comment += f'⚠️ **Warning: Coverage decreased by {ext_diff:.2f}%**\n\n'
+        comment += 'Consider adding tests to cover your changes.\n\n'
+    else:
+        comment += '✅ Coverage increased or remained the same\n\n'
+
+    # Webview coverage
+    comment += '### Webview Coverage\n\n'
+    comment += f'Base branch: {base_web_cov:.0f}%\n\n'
+    comment += f'PR branch: {pr_web_cov:.0f}%\n\n'
+
+    if web_decreased:
+        comment += f'⚠️ **Warning: Coverage decreased by {web_diff:.2f}%**\n\n'
+        comment += 'Consider adding tests to cover your changes.\n\n'
+    else:
+        comment += '✅ Coverage increased or remained the same\n\n'
+
+    # Overall assessment
+    comment += '### Overall Assessment\n\n'
+    if ext_decreased or web_decreased:
+        comment += '⚠️ **Test coverage has decreased in this PR**\n\n'
+        comment += 'Please consider adding tests to maintain or improve coverage.\n\n'
+    else:
+        comment += '✅ **Test coverage has been maintained or improved**\n\n'
+
+    # Add timestamp
+    comment += f'\n\n<sub>Last updated: {datetime.now().isoformat()}</sub>'
+    
+    return comment
+
+def post_comment(comment_path, pr_number, repo, token=None):
+    """
+    Post a comment to a GitHub PR.
+    
+    Args:
+        comment_path: Path to the file containing the comment text
+        pr_number: PR number
+        repo: Repository in the format "owner/repo"
+        token: GitHub token
+    """
+    if not file_exists(comment_path):
+        log(f"Error: Comment file {comment_path} does not exist")
+        return
+    
+    with open(comment_path, 'r') as f:
+        comment_body = f.read()
+    
+    if not token:
+        token = os.environ.get('GITHUB_TOKEN')
+        if not token:
+            log("Error: GitHub token not provided")
+            return
+    
+    # Find existing comment
+    headers = {
+        'Authorization': f'token {token}',
+        'Accept': 'application/vnd.github.v3+json'
+    }
+    
+    # Get all comments
+    comments_url = f'https://api.github.com/repos/{repo}/issues/{pr_number}/comments'
+    log(f"Getting comments from: {comments_url}")
+    response = requests.get(comments_url, headers=headers)
+    
+    if response.status_code != 200:
+        log(f"Error getting comments: {response.status_code} - {response.text}")
+        return
+    
+    comments = response.json()
+    log(f"Found {len(comments)} existing comments")
+    
+    # Find comment with our identifier
+    comment_id = None
+    for comment in comments:
+        if '<!-- COVERAGE_REPORT -->' in comment['body']:
+            comment_id = comment['id']
+            log(f"Found existing coverage report comment with ID: {comment_id}")
+            break
+    
+    if comment_id:
+        # Update existing comment
+        update_url = f'https://api.github.com/repos/{repo}/issues/comments/{comment_id}'
+        log(f"Updating existing comment at: {update_url}")
+        response = requests.patch(update_url, headers=headers, json={'body': comment_body})
+        
+        if response.status_code == 200:
+            log(f"Successfully updated existing comment: {comment_id}")
+        else:
+            log(f"Error updating comment: {response.status_code} - {response.text}")
+    else:
+        # Create new comment
+        log(f"Creating new comment at: {comments_url}")
+        response = requests.post(comments_url, headers=headers, json={'body': comment_body})
+        
+        if response.status_code == 201:
+            log("Successfully created new comment")
+        else:
+            log(f"Error creating comment: {response.status_code} - {response.text}")
+
+def set_github_output(name, value):
+    """
+    Set GitHub Actions output variable.
+    
+    Args:
+        name: Output variable name
+        value: Output variable value
+    """
+    # Write to the GitHub output file if available
+    if 'GITHUB_OUTPUT' in os.environ:
+        with open(os.environ['GITHUB_OUTPUT'], 'a') as f:
+            f.write(f"{name}={value}\n")
+    else:
+        # Fallback to the deprecated method for backward compatibility
+        log(f"::set-output name={name}::{value}")
+    
+    # Also print for human readability
+    log(f"{name}: {value}")

.github/scripts/coverage_check/util.py

@@ -0,0 +1,245 @@
+"""
+Utility module.
+This module provides utility functions used across the coverage check scripts.
+"""
+
+import os
+import sys
+import re
+import shlex
+import subprocess
+import traceback
+from typing import List, Tuple, Dict, Any, Optional, Union
+
+# List of allowed commands and their arguments
+ALLOWED_COMMANDS = {
+    'xvfb-run': ['-a'],
+    'npm': ['run', 'test:coverage', 'ci', 'install', '--no-save', '@vitest/coverage-v8', 'check-types', 'lint', 'format', 'compile'],
+    'cd': ['webview-ui'],
+    'python': ['-m', 'coverage_check'],
+    'git': ['fetch', 'checkout', 'origin'],
+}
+
+def is_safe_command(command: Union[str, List[str]]) -> bool:
+    """
+    Check if a command is safe to execute.
+    
+    Args:
+        command: Command to check (string or list)
+        
+    Returns:
+        True if command is safe, False otherwise
+    """
+    # Convert string command to list
+    if isinstance(command, str):
+        try:
+            cmd_parts = shlex.split(command)
+        except ValueError:
+            return False
+    else:
+        cmd_parts = command
+
+    if not cmd_parts:
+        return False
+
+    # Get base command
+    base_cmd = os.path.basename(cmd_parts[0])
+    
+    # Check if command is in allowed list
+    if base_cmd not in ALLOWED_COMMANDS:
+        return False
+        
+    # For each argument, check for suspicious patterns
+    for arg in cmd_parts[1:]:
+        # Check for shell metacharacters
+        if re.search(r'[;&|`$]', arg):
+            return False
+        # Check for path traversal
+        if '..' in arg and not (base_cmd == 'npm' and arg.startswith('@')):
+            return False
+            
+    return True
+
+def log(message: str) -> None:
+    """
+    Write a message to stdout and flush.
+    
+    Args:
+        message: The message to write
+    """
+    sys.stdout.write(f"{message}\n")
+    sys.stdout.flush()
+
+def file_exists(file_path: str) -> bool:
+    """
+    Check if a file exists.
+    
+    Args:
+        file_path: Path to the file
+        
+    Returns:
+        True if the file exists, False otherwise
+    """
+    return os.path.exists(file_path) and os.path.isfile(file_path)
+
+def get_file_size(file_path: str) -> int:
+    """
+    Get the size of a file in bytes.
+    
+    Args:
+        file_path: Path to the file
+        
+    Returns:
+        Size of the file in bytes, or 0 if the file doesn't exist
+    """
+    if file_exists(file_path):
+        return os.path.getsize(file_path)
+    return 0
+
+def list_directory(dir_path: str) -> List[Tuple[str, Union[int, str]]]:
+    """
+    List the contents of a directory.
+    
+    Args:
+        dir_path: Path to the directory
+        
+    Returns:
+        List of (name, size) tuples for each file/directory in the directory
+    """
+    if not os.path.exists(dir_path) or not os.path.isdir(dir_path):
+        return []
+    
+    contents = []
+    for item in os.listdir(dir_path):
+        item_path = os.path.join(dir_path, item)
+        if os.path.isfile(item_path):
+            contents.append((item, os.path.getsize(item_path)))
+        else:
+            contents.append((item, "DIR"))
+    
+    return contents
+
+def read_file_content(file_path: str, default: str = "") -> str:
+    """
+    Read file content with error handling.
+    
+    Args:
+        file_path: Path to the file
+        default: Default value to return if file cannot be read
+        
+    Returns:
+        File content or default value
+    """
+    if not file_exists(file_path):
+        log(f"File does not exist: {file_path}")
+        return default
+    
+    try:
+        with open(file_path, 'r') as f:
+            return f.read()
+    except Exception as e:
+        log(f"Error reading file {file_path}: {e}")
+        return default
+
+def write_file_content(file_path: str, content: str) -> bool:
+    """
+    Write content to file with error handling.
+    
+    Args:
+        file_path: Path to the file
+        content: Content to write
+        
+    Returns:
+        True if successful, False otherwise
+    """
+    try:
+        # Create directory if it doesn't exist
+        os.makedirs(os.path.dirname(file_path), exist_ok=True)
+        
+        with open(file_path, 'w') as f:
+            f.write(content)
+        return True
+    except Exception as e:
+        log(f"Error writing to file {file_path}: {e}")
+        return False
+
+def run_command(command: Union[str, List[str]], capture_output: bool = True) -> Tuple[int, str, str]:
+    """
+    Run a command and return the result.
+    
+    Args:
+        command: Command to run (string or list)
+        capture_output: Whether to capture stdout/stderr
+        
+    Returns:
+        Tuple of (returncode, stdout, stderr)
+    """
+    if not is_safe_command(command):
+        error_msg = f"Unsafe command detected: {command}"
+        log(error_msg)
+        return 1, "", error_msg
+        
+    log(f"Running command: {command}")
+    try:
+        # Convert string command to list
+        if isinstance(command, str):
+            cmd_list = shlex.split(command)
+        else:
+            cmd_list = command
+            
+        result = subprocess.run(
+            cmd_list,
+            shell=False,  # Never use shell=True for security
+            capture_output=capture_output,
+            text=True
+        )
+        log(f"Command exit code: {result.returncode}")
+        return result.returncode, result.stdout, result.stderr
+    except Exception as e:
+        log(f"Error running command: {e}")
+        log(traceback.format_exc())
+        return 1, "", str(e)
+
+def find_pattern(content: str, pattern: str, group: int = 0, 
+                default: Optional[str] = None) -> Optional[str]:
+    """
+    Find a pattern in content and return the specified group.
+    
+    Args:
+        content: Text content to search
+        pattern: Regex pattern to search for
+        group: Group number to return (default: 0 for entire match)
+        default: Default value to return if pattern not found
+        
+    Returns:
+        Matched text or default value
+    """
+    match = re.search(pattern, content, re.DOTALL)
+    if match:
+        return match.group(group)
+    return default
+
+def get_env_var(name: str, default: Optional[str] = None) -> Optional[str]:
+    """
+    Get environment variable with default value.
+    
+    Args:
+        name: Environment variable name
+        default: Default value if not set
+        
+    Returns:
+        Environment variable value or default
+    """
+    return os.environ.get(name, default)
+
+def format_exception(e: Exception) -> str:
+    """
+    Format an exception with traceback for logging.
+    
+    Args:
+        e: Exception to format
+        
+    Returns:
+        Formatted exception string
+    """
+    return f"{type(e).__name__}: {str(e)}\n{traceback.format_exc()}"

.github/scripts/coverage_check/workflow.py

@@ -0,0 +1,432 @@
+"""
+Workflow module.
+This module handles the main workflow logic for running coverage tests and processing results.
+"""
+
+import os
+import re
+import sys
+import subprocess
+import traceback
+
+from .extraction import run_coverage, compare_coverage, extract_coverage
+from .github_api import generate_comment, post_comment, set_github_output
+from .util import log, file_exists, get_file_size, list_directory, run_command
+
+def is_valid_branch_name(branch_name: str) -> bool:
+    """
+    Validate a git branch name.
+    
+    Args:
+        branch_name: Branch name to validate
+        
+    Returns:
+        True if valid, False otherwise
+    """
+    # Check for common branch name patterns
+    if not re.match(r'^[a-zA-Z0-9_\-./]+$', branch_name):
+        return False
+    
+    # Check for path traversal
+    if '..' in branch_name:
+        return False
+        
+    # Check for shell metacharacters
+    if re.search(r'[;&|`$]', branch_name):
+        return False
+        
+    return True
+
+def checkout_branch(branch_name: str) -> None:
+    """
+    Checkout a branch for testing.
+    
+    Args:
+        branch_name: Branch name to checkout
+        
+    Raises:
+        RuntimeError: If branch checkout fails
+        ValueError: If branch name is invalid
+    """
+    if not is_valid_branch_name(branch_name):
+        raise ValueError(f"Invalid branch name: {branch_name}")
+        
+    log(f"=== Checking out branch: {branch_name} ===")
+    
+    # Fetch the branch
+    returncode, stdout, stderr = run_command(['git', 'fetch', 'origin', branch_name])
+    if returncode != 0:
+        log(f"ERROR: Failed to fetch branch {branch_name}")
+        log(f"Error details: {stderr}")
+        raise RuntimeError(f"Git fetch failed: {stderr}")
+    
+    # Checkout the branch
+    returncode, stdout, stderr = run_command(['git', 'checkout', branch_name])
+    if returncode != 0:
+        log(f"ERROR: Failed to checkout branch {branch_name}")
+        log(f"Error details: {stderr}")
+        raise RuntimeError(f"Git checkout failed: {stderr}")
+    
+    log(f"Successfully checked out branch: {branch_name}")
+
+def extract_extension_coverage_from_file(file_path):
+    """Extract extension coverage from file when run_coverage returns 0."""
+    if not file_exists(file_path):
+        log(f"File {file_path} does not exist, cannot extract extension coverage")
+        return 0.0
+    
+    file_size = get_file_size(file_path)
+    if file_size == 0:
+        log(f"File {file_path} is empty, cannot extract extension coverage")
+        return 0.0
+        
+    log(f"Extension coverage is 0.0, trying to read from file directly: {file_path} (size: {file_size} bytes)")
+    with open(file_path, 'r') as f:
+        content = f.read()
+        # Extract the percentage from the "Lines" row in the coverage summary
+        # Pattern: Lines : xx.xx% ( xxxxxxx/xxxxxxx )
+        lines_match = re.search(r'Lines\s*:\s*(\d+\.\d+)%', content)
+        if lines_match:
+            coverage = float(lines_match.group(1))
+            log(f"Found extension coverage in file: {coverage}%")
+            return coverage
+    return 0.0
+
+def extract_webview_coverage_from_file(file_path):
+    """Extract webview coverage from file when run_coverage returns 0."""
+    if not file_exists(file_path):
+        log(f"File {file_path} does not exist, cannot extract webview coverage")
+        return 0.0
+    
+    file_size = get_file_size(file_path)
+    if file_size == 0:
+        log(f"File {file_path} is empty, cannot extract webview coverage")
+        return 0.0
+        
+    log(f"Webview coverage is 0.0, trying to read from file directly: {file_path} (size: {file_size} bytes)")
+    with open(file_path, 'r') as f:
+        content = f.read()
+        # Extract the percentage from the "% Lines" column in the "All files" row
+        # Pattern: All files | xx.xx | xx.xx | xx.xx | xx.xx |
+        all_files_match = re.search(r'All files\s+\|\s+\d+\.\d+\s+\|\s+\d+\.\d+\s+\|\s+\d+\.\d+\s+\|\s+(\d+\.\d+)', content)
+        if all_files_match:
+            coverage = float(all_files_match.group(1))
+            log(f"Found webview coverage in file: {coverage}%")
+            return coverage
+    return 0.0
+
+def run_extension_coverage(branch_name=None):
+    """Run extension coverage tests and extract results."""
+    prefix = 'base_' if branch_name else ''
+    file_path = f"{prefix}extension_coverage.txt"
+    
+    # Run coverage tests
+    ext_cov = run_coverage(
+        ["xvfb-run", "-a", "npm", "run", "test:coverage"], 
+        file_path, 
+        "extension"
+    )
+    
+    # If coverage is 0.0, try to extract from file directly
+    if ext_cov == 0.0:
+        ext_cov = extract_extension_coverage_from_file(file_path)
+        
+    return ext_cov
+
+def run_webview_coverage(branch_name=None):
+    """Run webview coverage tests and extract results."""
+    prefix = 'base_' if branch_name else ''
+    file_path = f"{prefix}webview_coverage.txt"
+    
+    # Save current directory
+    original_dir = os.getcwd()
+    
+    try:
+        # Change to webview-ui directory
+        os.chdir('webview-ui')
+        
+        # Install coverage dependency
+        returncode, stdout, stderr = run_command(["npm", "install", "--no-save", "@vitest/coverage-v8"])
+        if returncode != 0:
+            log(f"Failed to install coverage dependency: {stderr}")
+            return 0.0
+        
+        # Run coverage tests from webview-ui directory
+        web_cov = run_coverage(
+            ["npm", "run", "test:coverage"],
+            os.path.join('..', file_path),
+            "webview"
+        )
+    finally:
+        # Always change back to original directory
+        os.chdir(original_dir)
+    
+    # If coverage is 0.0, try to extract from file directly
+    if web_cov == 0.0:
+        web_cov = extract_webview_coverage_from_file(file_path)
+        
+    return web_cov
+
+def run_branch_coverage(branch_name=None):
+    """
+    Run coverage tests for a branch.
+    
+    Args:
+        branch_name: Name of the branch to checkout before running tests (optional)
+        
+    Returns:
+        Tuple of (extension_coverage, webview_coverage)
+    """
+    # Checkout branch if specified
+    if branch_name:
+        checkout_branch(branch_name)
+    
+    # Run coverage tests
+    log(f"=== Running coverage tests{' for ' + branch_name if branch_name else ''} ===")
+    
+    # Run extension and webview coverage
+    ext_cov = run_extension_coverage(branch_name)
+    web_cov = run_webview_coverage(branch_name)
+    
+    return ext_cov, web_cov
+
+def find_potential_coverage_files():
+    """Find potential coverage files in the current directory and webview-ui."""
+    log("Searching for potential coverage files...")
+    
+    # Find files in current directory
+    current_dir_files = list_directory('.')
+    for name, size in current_dir_files:
+        if 'coverage' in name.lower() and size != "DIR":
+            log(f"Found potential coverage file: {name} (size: {size} bytes)")
+    
+    # Find files in webview-ui directory
+    if os.path.exists('webview-ui') and os.path.isdir('webview-ui'):
+        webview_files = list_directory('webview-ui')
+        for name, size in webview_files:
+            if 'coverage' in name.lower() and size != "DIR":
+                log(f"Found potential webview coverage file: webview-ui/{name} (size: {size} bytes)")
+    else:
+        log("webview-ui directory not found")
+
+def generate_warnings(base_ext_cov, pr_ext_cov, ext_decreased, ext_diff, 
+                     base_web_cov, pr_web_cov, web_decreased, web_diff):
+    """Generate warnings for coverage decreases."""
+    if not (ext_decreased or web_decreased):
+        return []
+        
+    warnings = [
+        "Test coverage has decreased in this PR",
+        f"Extension coverage: {base_ext_cov}% -> {pr_ext_cov}% (Diff: {ext_diff}%)",
+        f"Webview coverage: {base_web_cov}% -> {pr_web_cov}% (Diff: {web_diff}%)"
+    ]
+    
+    # Additional warning for significant decrease (more than 1%)
+    if ext_decreased and ext_diff > 1.0:
+        warnings.append(f"Extension coverage decreased by more than 1% ({ext_diff}%). Consider adding tests to cover your changes.")
+    
+    if web_decreased and web_diff > 1.0:
+        warnings.append(f"Webview coverage decreased by more than 1% ({web_diff}%). Consider adding tests to cover your changes.")
+        
+    return warnings
+
+def output_warnings(warnings):
+    """Output warnings to GitHub step summary and console."""
+    if not warnings:
+        return
+        
+    # Get the GitHub step summary file path from environment variable
+    github_step_summary = os.environ.get('GITHUB_STEP_SUMMARY')
+    
+    # Write to GitHub step summary if available
+    if github_step_summary:
+        with open(github_step_summary, 'a') as f:
+            f.write("## Coverage Warnings\n\n")
+            for warning in warnings:
+                f.write(f"⚠️ {warning}\n\n")
+    
+    # Also output to console with ::warning:: syntax for backward compatibility
+    for warning in warnings:
+        log(f"::warning::{warning}")
+
+def output_github_results(pr_ext_cov, pr_web_cov, base_ext_cov, base_web_cov, 
+                         ext_decreased, ext_diff, web_decreased, web_diff):
+    """Output results for GitHub Actions."""
+    set_github_output("pr_extension_coverage", pr_ext_cov)
+    set_github_output("pr_webview_coverage", pr_web_cov)
+    set_github_output("base_extension_coverage", base_ext_cov)
+    set_github_output("base_webview_coverage", base_web_cov)
+    set_github_output("extension_decreased", str(ext_decreased).lower())
+    set_github_output("extension_diff", ext_diff)
+    set_github_output("webview_decreased", str(web_decreased).lower())
+    set_github_output("webview_diff", web_diff)
+
+def extract_pr_coverage_from_artifacts():
+    """
+    Extract PR branch coverage from artifact files.
+    
+    Returns:
+        Tuple of (extension_coverage, webview_coverage)
+        
+    Raises:
+        SystemExit: If the coverage files don't exist
+    """
+    log("=== Extracting PR branch coverage from artifacts ===")
+    
+    # Check if the coverage files exist
+    ext_file_path = "extension_coverage.txt"
+    web_file_path = "webview-ui/webview_coverage.txt"
+    
+    # Extract extension coverage
+    log(f"Extracting extension coverage from {ext_file_path}")
+    if not file_exists(ext_file_path):
+        error_msg = f"ERROR: PR extension coverage file {ext_file_path} not found"
+        log(error_msg)
+        
+        # List directory contents for debugging
+        log("Current directory contents:")
+        try:
+            dir_contents = list_directory('.')
+            for name, size in dir_contents:
+                log(f"  {name} - {size}\n")
+        except Exception as e:
+            log(f"Error listing directory: {e}")
+        
+        sys.exit(1)  # Exit with error code to fail the workflow
+    
+    ext_cov = extract_extension_coverage_from_file(ext_file_path)
+    log(f"PR extension coverage from artifact: {ext_cov}%")
+    
+    # Extract webview coverage
+    log(f"Extracting webview coverage from {web_file_path}")
+    if not file_exists(web_file_path):
+        error_msg = f"ERROR: PR webview coverage file {web_file_path} not found"
+        log(error_msg)
+        
+        # Check if the webview-ui directory exists
+        if not os.path.exists('webview-ui'):
+            log("ERROR: webview-ui directory not found")
+        else:
+            # List webview-ui directory contents for debugging
+            log("webview-ui directory contents:")
+            try:
+                dir_contents = list_directory('webview-ui')
+                for name, size in dir_contents:
+                    log(f"  {name} - {size}")
+            except Exception as e:
+                log(f"Error listing directory: {e}")
+        
+        sys.exit(1)  # Exit with error code to fail the workflow
+    
+    web_cov = extract_webview_coverage_from_file(web_file_path)
+    log(f"PR webview coverage from artifact: {web_cov}%")
+    
+    return ext_cov, web_cov
+
+def process_coverage_workflow(args):
+    """
+    Process the entire coverage workflow.
+    
+    Args:
+        args: Command line arguments
+    """
+    # Initialize all variables at the start
+    pr_ext_cov = 0.0
+    pr_web_cov = 0.0
+    base_ext_cov = 0.0
+    base_web_cov = 0.0
+    ext_decreased = False
+    ext_diff = 0.0
+    web_decreased = False
+    web_diff = 0.0
+    
+    try:
+        # Validate branch name
+        if not is_valid_branch_name(args.base_branch):
+            raise ValueError(f"Invalid base branch name: {args.base_branch}")
+        
+        # Check if we're running in GitHub Actions
+        is_github_actions = 'GITHUB_ACTIONS' in os.environ
+        if is_github_actions:
+            log("Running in GitHub Actions environment")
+        
+        # Extract PR branch coverage from artifacts (from test job)
+        pr_ext_cov, pr_web_cov = extract_pr_coverage_from_artifacts()
+        
+        # Verify PR coverage values
+        if pr_ext_cov == 0.0:
+            log("WARNING: PR extension coverage is 0.0, this may indicate an issue with the coverage report")
+            find_potential_coverage_files()
+        
+        if pr_web_cov == 0.0:
+            log("WARNING: PR webview coverage is 0.0, this may indicate an issue with the coverage report")
+            find_potential_coverage_files()
+        
+        # Run base branch coverage
+        log(f"=== Running base branch coverage for {args.base_branch} ===")
+        base_ext_cov, base_web_cov = run_branch_coverage(args.base_branch)
+        
+        # Verify base coverage values
+        if base_ext_cov == 0.0:
+            log("WARNING: Base extension coverage is 0.0, this may indicate an issue with the coverage report")
+        
+        if base_web_cov == 0.0:
+            log("WARNING: Base webview coverage is 0.0, this may indicate an issue with the coverage report")
+        
+        # Compare coverage
+        log("=== Comparing extension coverage ===")
+        ext_decreased, ext_diff = compare_coverage(base_ext_cov, pr_ext_cov)
+        
+        log("=== Comparing webview coverage ===")
+        web_decreased, web_diff = compare_coverage(base_web_cov, pr_web_cov)
+        
+        # Print summary of coverage values
+        log("\n=== Coverage Summary ===")
+        log(f"PR extension coverage: {pr_ext_cov}%")
+        log(f"Base extension coverage: {base_ext_cov}%")
+        log(f"Extension coverage change: {'+' if not ext_decreased else '-'}{ext_diff}%")
+        log(f"PR webview coverage: {pr_web_cov}%")
+        log(f"Base webview coverage: {base_web_cov}%")
+        log(f"Webview coverage change: {'+' if not web_decreased else '-'}{web_diff}%")
+        
+        # Generate and output warnings
+        warnings = generate_warnings(
+            base_ext_cov, pr_ext_cov, ext_decreased, ext_diff,
+            base_web_cov, pr_web_cov, web_decreased, web_diff
+        )
+        output_warnings(warnings)
+        
+        # Generate comment
+        log("=== Generating comment ===")
+        comment = generate_comment(
+            base_ext_cov, pr_ext_cov, str(ext_decreased).lower(), ext_diff,
+            base_web_cov, pr_web_cov, str(web_decreased).lower(), web_diff
+        )
+        
+        # Save comment to file
+        with open("coverage_comment.md", "w") as f:
+            f.write(comment)
+        
+        # Post comment if PR number is provided
+        if args.pr_number:
+            log(f"=== Posting comment to PR #{args.pr_number} ===")
+            post_comment("coverage_comment.md", args.pr_number, args.repo, args.token)
+        
+        # Output results for GitHub Actions
+        output_github_results(
+            pr_ext_cov, pr_web_cov, base_ext_cov, base_web_cov,
+            ext_decreased, ext_diff, web_decreased, web_diff
+        )
+        
+    except Exception as e:
+        log(f"ERROR in process_coverage_workflow: {e}")
+        traceback.print_exc()
+        
+        # Try to output results even if there was an error
+        try:
+            output_github_results(
+                pr_ext_cov, pr_web_cov, base_ext_cov, base_web_cov,
+                ext_decreased, ext_diff, web_decreased, web_diff
+            )
+        except Exception as e2:
+            log(f"ERROR outputting GitHub results: {e2}")

.github/scripts/overwrite_changeset_changelog.py

@@ -0,0 +1,79 @@
+"""
+This script updates a specific version's release notes section in CHANGELOG.md with new content
+or reformats existing content.
+
+The script:
+1. Takes a version number, changelog path, and optionally new content as input from environment variables
+2. Finds the section in the changelog for the specified version
+3. Either:
+   a) Replaces the content with new content if provided, or
+   b) Reformats existing content by:
+      - Removing the first two lines of the changeset format
+      - Ensuring version numbers are wrapped in square brackets
+4. Writes the updated changelog back to the file
+
+Environment Variables:
+    CHANGELOG_PATH: Path to the changelog file (defaults to 'CHANGELOG.md')
+    VERSION: The version number to update/format
+    PREV_VERSION: The previous version number (used to locate section boundaries)
+    NEW_CONTENT: Optional new content to insert for this version
+"""
+
+#!/usr/bin/env python3
+
+import os
+
+CHANGELOG_PATH = os.environ.get("CHANGELOG_PATH", "CHANGELOG.md")
+VERSION = os.environ['VERSION']
+PREV_VERSION = os.environ.get("PREV_VERSION", "")
+NEW_CONTENT = os.environ.get("NEW_CONTENT", "")
+
+def overwrite_changelog_section(changelog_text: str, new_content: str):
+    # Find the section for the specified version
+    version_pattern = f"## {VERSION}\n"
+    unformmatted_prev_version_pattern = f"## {PREV_VERSION}\n"
+    prev_version_pattern = f"## [{PREV_VERSION}]\n"
+    print(f"latest version: {VERSION}")
+    print(f"prev_version: {PREV_VERSION}")
+
+    notes_start_index = changelog_text.find(version_pattern) + len(version_pattern)
+    notes_end_index = changelog_text.find(prev_version_pattern, notes_start_index) if PREV_VERSION and (prev_version_pattern in changelog_text or unformmatted_prev_version_pattern in changelog_text) else len(changelog_text)
+
+    if new_content:
+        return changelog_text[:notes_start_index] + f"{new_content}\n" + changelog_text[notes_end_index:]
+    else:
+        changeset_lines = changelog_text[notes_start_index:notes_end_index].split("\n")
+        filtered_lines = []
+        for line in changeset_lines:
+            # If the previous line is a changeset format
+            if len(filtered_lines) > 1 and filtered_lines[-1].startswith("### "):
+                # Remove the last two lines from the filted_lines
+                filtered_lines.pop()
+                filtered_lines.pop()
+            else:
+                filtered_lines.append(line.strip())
+
+        # Prepend a new line to the first line of filtered_lines
+        if filtered_lines:
+            filtered_lines[0] = "\n" + filtered_lines[0]
+
+        # Print filted_lines wiht a "\n" at the end of each line
+        for line in filtered_lines:
+            print(line.strip())
+
+        parsed_lines = "\n".join(line for line in filtered_lines)
+        updated_changelog = changelog_text[:notes_start_index] + parsed_lines + changelog_text[notes_end_index:]
+        return updated_changelog
+
+with open(CHANGELOG_PATH, 'r') as f:
+    changelog_content = f.read()
+
+new_changelog = overwrite_changelog_section(changelog_content, NEW_CONTENT)
+# print("----------------------------------------------------------------------------------")
+# print(new_changelog)
+# print("----------------------------------------------------------------------------------")
+# Write back to CHANGELOG.md
+with open(CHANGELOG_PATH, 'w') as f:
+    f.write(new_changelog)
+
+print(f"{CHANGELOG_PATH} updated successfully!")

.github/scripts/tests/coverage_check_test.py

@@ -0,0 +1,282 @@
+#!/usr/bin/env python3
+"""
+Tests for coverage_check script.
+"""
+
+import os
+import sys
+import unittest
+import subprocess
+import tempfile
+from unittest.mock import patch, MagicMock, call, mock_open
+
+# Add parent directory to path so we can import coverage modules
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
+from coverage_check import extract_coverage, compare_coverage, set_verbose, generate_comment, post_comment, set_github_output
+from coverage_check.util import log, file_exists, get_file_size, list_directory
+
+
+class TestCoverage(unittest.TestCase):
+    # Class variables to store coverage files
+    temp_dir = None
+    extension_coverage_file = None
+    webview_coverage_file = None
+    
+    @classmethod
+    def setUpClass(cls):
+        """Set up test environment once for all tests."""
+        # Create temporary directory for test files
+        cls.temp_dir = tempfile.TemporaryDirectory()
+        cls.extension_coverage_file = os.path.join(cls.temp_dir.name, 'extension_coverage.txt')
+        cls.webview_coverage_file = os.path.join(cls.temp_dir.name, 'webview_coverage.txt')
+        
+        # Run actual tests to generate coverage reports
+        cls.generate_coverage_reports()
+        
+        # Verify files exist and are not empty
+        assert os.path.exists(cls.extension_coverage_file), \
+            f"Extension coverage file {cls.extension_coverage_file} does not exist"
+        assert os.path.getsize(cls.extension_coverage_file) > 0, \
+            f"Extension coverage file {cls.extension_coverage_file} is empty"
+        assert os.path.exists(cls.webview_coverage_file), \
+            f"Webview coverage file {cls.webview_coverage_file} does not exist"
+        assert os.path.getsize(cls.webview_coverage_file) > 0, \
+            f"Webview coverage file {cls.webview_coverage_file} is empty"
+
+    @classmethod
+    def tearDownClass(cls):
+        """Clean up test environment after all tests."""
+        if cls.temp_dir:
+            cls.temp_dir.cleanup()
+
+    @classmethod
+    def generate_coverage_reports(cls):
+        """Generate real coverage reports by running tests."""
+        log("Generating coverage reports (this may take a while)...")
+        
+        # Run extension tests with coverage
+        try:
+            # Get absolute paths
+            root_dir = os.path.abspath(os.path.join(os.path.dirname(__file__), '../../..'))
+            webview_dir = os.path.join(root_dir, 'webview-ui')
+            
+            # Use xvfb-run on Linux
+            if sys.platform.startswith('linux'):
+                cmd = f"cd {root_dir} && xvfb-run -a npm run test:coverage > {cls.extension_coverage_file} 2>&1"
+            else:
+                cmd = f"cd {root_dir} && npm run test:coverage > {cls.extension_coverage_file} 2>&1"
+            
+            log("Running extension tests...")
+            log(f"Command: {cmd}")
+            result = subprocess.run(cmd, shell=True, check=False, capture_output=True, text=True)
+            log(f"Extension tests exit code: {result.returncode}")
+            
+            # Run webview tests with coverage
+            log("Running webview tests...")
+            cmd = f"cd {webview_dir} && npm run test:coverage > {cls.webview_coverage_file} 2>&1"
+            log(f"Command: {cmd}")
+            result = subprocess.run(cmd, shell=True, check=False, capture_output=True, text=True)
+            log(f"Webview tests exit code: {result.returncode}")
+            
+            # Verify files were created
+            if file_exists(cls.extension_coverage_file):
+                ext_size = get_file_size(cls.extension_coverage_file)
+                log(f"Extension coverage file created: {cls.extension_coverage_file} (size: {ext_size} bytes)")
+            else:
+                log(f"WARNING: Extension coverage file was not created: {cls.extension_coverage_file}")
+                
+            if file_exists(cls.webview_coverage_file):
+                web_size = get_file_size(cls.webview_coverage_file)
+                log(f"Webview coverage file created: {cls.webview_coverage_file} (size: {web_size} bytes)")
+            else:
+                log(f"WARNING: Webview coverage file was not created: {cls.webview_coverage_file}")
+            
+            log("Coverage reports generation completed.")
+        except Exception as e:
+            log(f"Error generating coverage reports: {e}")
+            import traceback
+            log(traceback.format_exc())
+            
+            # Create empty files if tests fail
+            log("Creating fallback coverage files...")
+            with open(cls.extension_coverage_file, 'w') as f:
+                f.write("No coverage data available")
+            with open(cls.webview_coverage_file, 'w') as f:
+                f.write("No coverage data available")
+
+    def test_extract_coverage(self):
+        """Test extract_coverage function with both extension and webview coverage."""
+        # Check if verbose mode is enabled
+        if '-v' in sys.argv or '--verbose' in sys.argv:
+            set_verbose(True)
+        
+        # Verify files exist before testing
+        self.assertTrue(file_exists(self.extension_coverage_file), 
+                       f"Extension coverage file does not exist: {self.extension_coverage_file}")
+        self.assertTrue(file_exists(self.webview_coverage_file), 
+                       f"Webview coverage file does not exist: {self.webview_coverage_file}")
+        
+        # Log file sizes
+        ext_size = get_file_size(self.extension_coverage_file)
+        web_size = get_file_size(self.webview_coverage_file)
+        log(f"Extension coverage file size: {ext_size} bytes")
+        log(f"Webview coverage file size: {web_size} bytes")
+        
+        # Test extension coverage
+        log("Testing extension coverage extraction...")
+        ext_coverage_pct = extract_coverage(self.extension_coverage_file, 'extension')
+        
+        # Check that coverage percentage is a float
+        self.assertIsInstance(ext_coverage_pct, float)
+        
+        # Check that coverage percentage is between 0 and 100
+        self.assertGreaterEqual(ext_coverage_pct, 0)
+        self.assertLessEqual(ext_coverage_pct, 100)
+        
+        # Log coverage percentage for debugging
+        log(f"Extension coverage: {ext_coverage_pct}%")
+        
+        # Test webview coverage
+        log("Testing webview coverage extraction...")
+        web_coverage_pct = extract_coverage(self.webview_coverage_file, 'webview')
+        
+        # Convert to float if it's an integer
+        if isinstance(web_coverage_pct, int):
+            web_coverage_pct = float(web_coverage_pct)
+        
+        # Check that coverage percentage is a float
+        self.assertIsInstance(web_coverage_pct, float)
+        
+        # Check that coverage percentage is between 0 and 100
+        self.assertGreaterEqual(web_coverage_pct, 0)
+        self.assertLessEqual(web_coverage_pct, 100)
+        
+        # Log coverage percentage for debugging
+        log(f"Webview coverage: {web_coverage_pct}%")
+
+    def test_compare_coverage(self):
+        """Test compare_coverage function."""
+        # Test with coverage increase
+        decreased, diff = compare_coverage(80, 90)
+        self.assertFalse(decreased)
+        self.assertEqual(diff, 10)
+        
+        # Test with coverage decrease
+        decreased, diff = compare_coverage(90, 80)
+        self.assertTrue(decreased)
+        self.assertEqual(diff, 10)
+        
+        # Test with no change
+        decreased, diff = compare_coverage(80, 80)
+        self.assertFalse(decreased)
+        self.assertEqual(diff, 0)
+
+    def test_generate_comment(self):
+        """Test generate_comment function."""
+        comment = generate_comment(
+            80, 90, 'false', 10,
+            70, 75, 'false', 5
+        )
+        
+        # Check that comment contains expected sections
+        self.assertIn('Coverage Report', comment)
+        self.assertIn('Extension Coverage', comment)
+        self.assertIn('Webview Coverage', comment)
+        self.assertIn('Overall Assessment', comment)
+        
+        # Check that comment contains coverage percentages
+        self.assertIn('Base branch: 80%', comment)
+        self.assertIn('PR branch: 90%', comment)
+        self.assertIn('Base branch: 70%', comment)
+        self.assertIn('PR branch: 75%', comment)
+        
+        # Check that comment contains correct assessment
+        self.assertIn('Coverage increased or remained the same', comment)
+        self.assertIn('Test coverage has been maintained or improved', comment)
+
+    @patch('coverage_check.requests.get')
+    @patch('coverage_check.requests.post')
+    @patch('coverage_check.requests.patch')
+    def test_post_comment_new(self, mock_patch, mock_post, mock_get):
+        """Test post_comment function when creating a new comment."""
+        # Create a temporary comment file
+        comment_file = os.path.join(self.temp_dir.name, 'comment.md')
+        with open(comment_file, 'w') as f:
+            f.write('<!-- COVERAGE_REPORT -->\nTest comment')
+        
+        # Mock the API responses
+        mock_get.return_value = MagicMock(status_code=200, json=lambda: [])
+        mock_post.return_value = MagicMock(status_code=201)
+        
+        # Test post_comment function
+        post_comment(comment_file, '123', 'owner/repo', 'token')
+        
+        # Check that the correct API calls were made
+        mock_get.assert_called_once()
+        mock_post.assert_called_once()
+        mock_patch.assert_not_called()
+
+    @patch('coverage_check.requests.get')
+    @patch('coverage_check.requests.post')
+    @patch('coverage_check.requests.patch')
+    def test_post_comment_update(self, mock_patch, mock_post, mock_get):
+        """Test post_comment function when updating an existing comment."""
+        # Create a temporary comment file
+        comment_file = os.path.join(self.temp_dir.name, 'comment.md')
+        with open(comment_file, 'w') as f:
+            f.write('<!-- COVERAGE_REPORT -->\nTest comment')
+        
+        # Mock the API responses
+        mock_get.return_value = MagicMock(
+            status_code=200, 
+            json=lambda: [{'id': 456, 'body': '<!-- COVERAGE_REPORT -->\nOld comment'}]
+        )
+        mock_patch.return_value = MagicMock(status_code=200)
+        
+        # Test post_comment function
+        post_comment(comment_file, '123', 'owner/repo', 'token')
+        
+        # Check that the correct API calls were made
+        mock_get.assert_called_once()
+        mock_patch.assert_called_once()
+        mock_post.assert_not_called()
+
+    def test_set_github_output(self):
+        """Test set_github_output function."""
+        # Capture stdout
+        with patch('sys.stdout', new=MagicMock()) as mock_stdout:
+            # Mock environment without GITHUB_OUTPUT
+            with patch.dict('os.environ', {}, clear=True):
+                set_github_output('test_name', 'test_value')
+                
+                # Check that the correct output was printed to stdout
+                mock_stdout.assert_has_calls([
+                    # GitHub Actions output format (deprecated method)
+                    call.write('::set-output name=test_name::test_value\n'),
+                    call.flush(),
+                    # Human readable format
+                    call.write('test_name: test_value\n'),
+                    call.flush()
+                ], any_order=False)
+                
+                # Reset mock for next test
+                mock_stdout.reset_mock()
+            
+            # Test with GITHUB_OUTPUT environment variable
+            with patch.dict('os.environ', {'GITHUB_OUTPUT': '/tmp/github_output'}), \
+                 patch('builtins.open', mock_open()) as mock_file:
+                set_github_output('test_name', 'test_value')
+                
+                # Check that file was written to
+                mock_file.assert_called_once_with('/tmp/github_output', 'a')
+                mock_file().write.assert_called_once_with('test_name=test_value\n')
+                
+                # Check that human readable output was printed
+                mock_stdout.assert_has_calls([
+                    call.write('test_name: test_value\n'),
+                    call.flush()
+                ], any_order=False)
+
+
+if __name__ == '__main__':
+    unittest.main()

.github/workflows/changeset-converter.yml

@@ -0,0 +1,112 @@
+name: Changeset Converter
+run-name: Changeset Conversion
+
+on:
+    workflow_dispatch:
+    pull_request:
+        types: [closed]
+
+env:
+    REPO_PATH: ${{ github.repository }}
+    GIT_REF: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || 'main' }}
+    NODE_VERSION: 20.18.1
+
+jobs:
+    # Job 1: Create version bump PR when changesets are merged to main
+    changeset-pr-version-bump:
+        if: |
+            github.event_name == 'workflow_dispatch' ||
+            (
+                github.event_name == 'pull_request' &&
+                github.event.pull_request.merged == true &&
+                github.event.pull_request.base.ref == 'main' &&
+                github.actor != 'github-actions'
+            )
+        runs-on: ubuntu-latest
+        permissions:
+            contents: write
+            pull-requests: write
+        steps:
+            - name: Check user for team affiliation
+              id: team_check
+              if: github.event_name == 'workflow_dispatch'
+              uses: morfien101/actions-authorized-user@4a3cfbf0bcb3cafe4a71710a278920c5d94bb38b
+              with:
+                  username: ${{ github.actor }}
+                  team: "deployer"
+                  github_token: ${{ secrets.GITHUB_TOKEN }}
+
+            - name: Check if user is authorized
+              if: github.event_name == 'workflow_dispatch'
+              run: |
+                  if [ "${{ steps.team_check.outputs.authorized }}" != "true" ]; then
+                      echo "User is not authorized to run this workflow."
+                      exit 1
+                  fi
+
+            - name: Git Checkout
+              uses: actions/checkout@v4
+              with:
+                  fetch-depth: 0
+                  ref: ${{ env.GIT_REF }}
+
+            - name: Setup Node.js
+              uses: actions/setup-node@v4
+              with:
+                  node-version: ${{ env.NODE_VERSION }}
+                  cache: "npm"
+
+            - name: Install Dependencies
+              run: npm install changeset
+
+            # Check if there are any new changesets to process
+            - name: Check for changesets
+              id: check-changesets
+              run: |
+                  NEW_CHANGESETS=$(find .changeset -name "*.md" ! -name "README.md" | wc -l | tr -d ' ')
+                  echo "Changesets diff with previous version: $NEW_CHANGESETS"
+                  echo "new_changesets=$NEW_CHANGESETS" >> $GITHUB_OUTPUT
+
+            # Create version bump PR using changesets/action if there are new changesets
+            - name: Create Changeset Pull Request
+              if: steps.check-changesets.outputs.new_changesets != '0'
+              uses: changesets/action@v1
+              with:
+                  commit: "changeset version bump"
+                  title: "Changeset version bump"
+                  version: npm run version-packages # This performs the changeset version bump
+              env:
+                  GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+            # Get current and previous versions to edit changelog entry
+            - name: Get version
+              id: get_version
+              run: |
+                  VERSION=$(git show HEAD:package.json | jq -r '.version')
+                  echo "version=$VERSION" >> $GITHUB_OUTPUT
+                  PREV_VERSION=$(git show origin/main:package.json | jq -r '.version')
+                  echo "prev_version=$PREV_VERSION" >> $GITHUB_OUTPUT
+                  echo "version=$VERSION"
+                  echo "prev_version=$PREV_VERSION"
+
+            # Update CHANGELOG.md with proper format
+            - name: Update Changelog Format
+              env:
+                  VERSION: ${{ steps.get_version.outputs.version }}
+                  PREV_VERSION: ${{ steps.get_version.outputs.prev_version }}
+              run: python .github/scripts/overwrite_changeset_changelog.py
+
+            # Commit and push changelog updates
+            - name: Push Changelog updates to Pull Request
+              run: |
+                  git config user.name "github-actions"
+                  git config user.email github-actions@github.com
+                  echo "Running git add and commit..."
+                  git add CHANGELOG.md
+                  git commit -m "Updating CHANGELOG.md format"
+                  git status
+                  echo "--------------------------------------------------------------------------------"
+                  echo "Pushing to remote..."
+                  echo "--------------------------------------------------------------------------------"
+                  CURRENT_BRANCH=$(git rev-parse --abbrev-ref HEAD)
+                  git push origin $CURRENT_BRANCH

.github/workflows/codespell.yml

@@ -0,0 +1,28 @@
+# Codespell configuration is within .codespellrc
+---
+name: Codespell
+
+on:
+    push:
+        branches: [main]
+    pull_request:
+        branches: [main]
+
+permissions:
+    contents: read
+
+jobs:
+    codespell:
+        if: false
+        name: Check for spelling errors
+        runs-on: ubuntu-latest
+
+        steps:
+            - name: Checkout
+              uses: actions/checkout@v4
+            - name: Annotate locations with typos
+              uses: codespell-project/codespell-problem-matcher@v1
+            - name: Codespell
+              uses: codespell-project/actions-codespell@v2
+              with:
+                  only_warn: 1

.github/workflows/publish.yml

@@ -70,8 +70,10 @@ jobs:
                   echo "version=$VERSION" >> $GITHUB_OUTPUT
 
             - name: Create Git Tag
+              id: create_tag
               run: |
                   VERSION=v${{ steps.get_version.outputs.version }}
+                  echo "tag=$VERSION" >> $GITHUB_OUTPUT
                   echo "Tagging with $VERSION"
                   git tag "$VERSION"
                   git push origin "$VERSION"
@@ -104,6 +106,7 @@ jobs:
             - name: Create GitHub Release
               uses: softprops/action-gh-release@v1
               with:
+                  tag_name: ${{ steps.create_tag.outputs.tag }}
                   files: "*.vsix"
                   # body: ${{ steps.changelog.outputs.content }}
                   generate_release_notes: true

.github/workflows/test.yml

@@ -23,7 +23,18 @@ jobs:
             - name: Setup Node.js environment
               uses: actions/setup-node@v4
               with:
-                  node-version: 20.15.1
+                  node-version: 22
+
+            # Setup Python for coverage script
+            - name: Setup Python
+              uses: actions/setup-python@v4
+              with:
+                  python-version: "3.10"
+
+            - name: Install Python dependencies
+              run: |
+                  python -m pip install --upgrade pip
+                  pip install requests
 
             # Cache root dependencies - only reuse if package-lock.json exactly matches
             - name: Cache root dependencies
@@ -58,5 +69,130 @@ jobs:
             - name: Prettier / Format Check
               run: npm run format
 
-            - name: Extension Tests
-              run: xvfb-run -a npm run test
+            # Build the extension before running tests
+            - name: Build Tests and Extension
+              run: npm run pretest
+
+            - name: Unit Tests
+              run: npm run test:unit
+
+            # Run extension tests with coverage
+            - name: Extension Tests with Coverage
+              id: extension_coverage
+              continue-on-error: true
+              run: |
+                  xvfb-run -a npm run test:coverage > extension_coverage.txt 2>&1
+                  PYTHONPATH=.github/scripts python -m coverage_check extract-coverage extension_coverage.txt --type=extension --github-output --verbose
+
+            # Run webview tests with coverage
+            - name: Webview Tests with Coverage
+              id: webview_coverage
+              continue-on-error: true
+              run: |
+                  cd webview-ui
+                  # Ensure coverage dependency is installed
+                  npm install --no-save @vitest/coverage-v8
+                  npm run test:coverage > webview_coverage.txt 2>&1 || true
+                  cd ..
+                  PYTHONPATH=.github/scripts python -m coverage_check extract-coverage webview-ui/webview_coverage.txt --type=webview --github-output --verbose
+
+            # Save coverage reports as artifacts (workflow-scoped)
+            - name: Save Coverage Reports
+              uses: actions/upload-artifact@v4
+              with:
+                  name: pr-coverage-reports
+                  path: |
+                      extension_coverage.txt
+                      webview-ui/webview_coverage.txt
+                  retention-period: workflow # Artifacts are automatically deleted when the workflow completes
+
+            # Set the check as failed if any of the tests failed
+            - name: Check for test failures
+              run: |
+                  # Check if any of the test steps failed
+                  # https://docs.github.com/en/actions/writing-workflows/choosing-what-your-workflow-does/accessing-contextual-information-about-workflow-runs#steps-context
+                  if [ "${{ steps.extension_coverage.outcome }}" != "success" ] || [ "${{ steps.webview_coverage.outcome }}" != "success" ]; then
+                      echo "Tests failed."
+                      cat extension_coverage.txt
+                      cat webview-ui/webview_coverage.txt
+                      exit 1
+                  fi
+
+    coverage:
+        needs: test
+        runs-on: ubuntu-latest
+        # Only run on PRs to main branch
+        if: github.event_name == 'pull_request' && github.base_ref == 'main'
+        steps:
+            - name: Checkout code
+              uses: actions/checkout@v4
+              with:
+                  fetch-depth: 0 # Fetch all history for accurate comparison
+
+            # Setup Python for coverage script
+            - name: Setup Python
+              uses: actions/setup-python@v4
+              with:
+                  python-version: "3.10"
+
+            - name: Install Python dependencies
+              run: |
+                  python -m pip install --upgrade pip
+                  pip install requests
+
+            - name: Setup Node.js environment
+              uses: actions/setup-node@v4
+              with:
+                  node-version: 22
+
+            # Cache root dependencies - only reuse if package-lock.json exactly matches
+            - name: Cache root dependencies
+              uses: actions/cache@v4
+              id: root-cache
+              with:
+                  path: node_modules
+                  key: ${{ runner.os }}-npm-${{ hashFiles('package-lock.json') }}
+
+            # Cache webview-ui dependencies - only reuse if package-lock.json exactly matches
+            - name: Cache webview-ui dependencies
+              uses: actions/cache@v4
+              id: webview-cache
+              with:
+                  path: webview-ui/node_modules
+                  key: ${{ runner.os }}-npm-webview-${{ hashFiles('webview-ui/package-lock.json') }}
+
+            - name: Install root dependencies
+              if: steps.root-cache.outputs.cache-hit != 'true'
+              run: npm ci
+
+            - name: Install webview-ui dependencies
+              if: steps.webview-cache.outputs.cache-hit != 'true'
+              run: cd webview-ui && npm ci
+
+            # Build the extension before running tests
+            - name: Build Extension
+              run: npm run compile
+
+            # Download coverage artifacts from test job
+            - name: Download Coverage Reports
+              uses: actions/download-artifact@v4
+              with:
+                  name: pr-coverage-reports
+                  path: . # Download to root directory to match expected paths
+
+            # Process coverage workflow
+            - name: Process coverage workflow
+              id: coverage
+              run: |
+                  # Extract PR number from GITHUB_REF
+                  PR_NUMBER=$(echo "$GITHUB_REF" | sed -e 's/refs\/pull\///' -e 's/\/merge//')
+
+                  # Run the coverage workflow from root directory
+                  PYTHONPATH=.github/scripts python -m coverage_check process-workflow \
+                    --base-branch ${{ github.base_ref }} \
+                    --pr-number $PR_NUMBER \
+                    --repo $GITHUB_REPOSITORY \
+                    --token ${{ secrets.GITHUB_TOKEN }} \
+                    --verbose
+              env:
+                  GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.gitignore

@@ -7,4 +7,15 @@ tmp
 
 .DS_Store
 
-pnpm-lock.yaml
+pnpm-lock.yaml
+
+.clineignore
+.venv
+.actrc
+
+# Ignore coverage directories and files
+coverage
+# But don't ignore the coverage scripts in .github/scripts/
+!.github/scripts/coverage/
+
+*evals.env

.mocharc.json

@@ -0,0 +1,6 @@
+{
+	"extension": ["ts"],
+	"spec": "src/**/__tests__/*.ts",
+	"require": ["ts-node/register", "source-map-support/register", "./src/test/requires.ts"],
+	"recursive": true
+}

.vscode-test.mjs

@@ -6,6 +6,10 @@ export default defineConfig({
 	mocha: {
 		ui: "bdd",
 		timeout: 20000, // Maximum time (in ms) that a test can run before failing
+		/** Set up alias path resolution during tests
+		 * @See {@link file://./test-setup.js}
+		 */
+		require: ["./test-setup.js"],
 	},
 	workspaceFolder: "test-workspace",
 	version: "stable",

.vscode/launch.json

@@ -9,9 +9,13 @@
 			"name": "Run Extension",
 			"type": "extensionHost",
 			"request": "launch",
-			"args": ["--extensionDevelopmentPath=${workspaceFolder}"],
+			"args": ["--extensionDevelopmentPath=${workspaceFolder}", "--disable-workspace-trust", "${workspaceFolder}"],
 			"outFiles": ["${workspaceFolder}/dist/**/*.js"],
-			"preLaunchTask": "${defaultBuildTask}"
+			"preLaunchTask": "${defaultBuildTask}",
+			"env": {
+				"IS_DEV": "true",
+				"DEV_WORKSPACE_FOLDER": "${workspaceFolder}"
+			}
 		}
 	]
 }

.vscode/tasks.json

@@ -4,16 +4,45 @@
 	"version": "2.0.0",
 	"tasks": [
 		{
-			"label": "watch",
-			"dependsOn": ["npm: build:webview", "npm: watch:tsc", "npm: watch:esbuild"],
+			"label": "npm: protos",
+			"type": "npm",
+			"script": "protos",
+			"problemMatcher": [],
+			"isBackground": false,
 			"presentation": {
-				"reveal": "never"
+				"reveal": "always"
+			},
+			"options": {
+				"env": {
+					"IS_DEV": "true"
+				}
+			}
+		},
+		{
+			"label": "watch",
+			"dependsOn": ["npm: protos", "npm: build:webview", "npm: dev:webview", "npm: watch:tsc", "npm: watch:esbuild"],
+			"presentation": {
+				"reveal": "always"
 			},
 			"group": {
 				"kind": "build",
 				"isDefault": true
 			}
 		},
+		{
+			"label": "watch:test",
+			"dependsOn": [
+				"npm: protos",
+				"npm: build:webview:test",
+				"npm: dev:webview",
+				"npm: watch:tsc",
+				"npm: watch:esbuild:test"
+			],
+			"presentation": {
+				"reveal": "always"
+			},
+			"group": "build"
+		},
 		{
 			"type": "npm",
 			"script": "build:webview",
@@ -21,9 +50,68 @@
 			"problemMatcher": [],
 			"isBackground": true,
 			"label": "npm: build:webview",
+			"dependsOn": ["npm: protos"],
 			"presentation": {
 				"group": "watch",
-				"reveal": "never"
+				"reveal": "always"
+			},
+			"options": {
+				"env": {
+					"IS_DEV": "true"
+				}
+			}
+		},
+		{
+			"type": "npm",
+			"script": "build:webview:test",
+			"group": "build",
+			"problemMatcher": [],
+			"isBackground": true,
+			"label": "npm: build:webview:test",
+			"dependsOn": ["npm: protos"],
+			"presentation": {
+				"group": "watch",
+				"reveal": "always"
+			},
+			"options": {
+				"env": {
+					"IS_DEV": "true",
+					"IS_TEST": "true"
+				}
+			}
+		},
+		{
+			"type": "npm",
+			"script": "dev:webview",
+			"group": "build",
+			"problemMatcher": [
+				{
+					"pattern": [
+						{
+							"regexp": ".",
+							"file": 1,
+							"location": 2,
+							"message": 3
+						}
+					],
+					"background": {
+						"activeOnStart": true,
+						"beginsPattern": ".",
+						"endsPattern": "."
+					}
+				}
+			],
+			"isBackground": true,
+			"label": "npm: dev:webview",
+			"dependsOn": ["npm: protos"],
+			"presentation": {
+				"group": "watch",
+				"reveal": "always"
+			},
+			"options": {
+				"env": {
+					"IS_DEV": "true"
+				}
 			}
 		},
 		{
@@ -33,9 +121,34 @@
 			"problemMatcher": "$esbuild-watch",
 			"isBackground": true,
 			"label": "npm: watch:esbuild",
+			"dependsOn": ["npm: protos"],
 			"presentation": {
 				"group": "watch",
-				"reveal": "never"
+				"reveal": "always"
+			},
+			"options": {
+				"env": {
+					"IS_DEV": "true"
+				}
+			}
+		},
+		{
+			"type": "npm",
+			"script": "watch:esbuild:test",
+			"group": "build",
+			"problemMatcher": "$esbuild-watch",
+			"isBackground": true,
+			"label": "npm: watch:esbuild:test",
+			"dependsOn": ["npm: protos"],
+			"presentation": {
+				"group": "watch",
+				"reveal": "always"
+			},
+			"options": {
+				"env": {
+					"IS_DEV": "true",
+					"IS_TEST": "true"
+				}
 			}
 		},
 		{
@@ -45,9 +158,10 @@
 			"problemMatcher": "$tsc-watch",
 			"isBackground": true,
 			"label": "npm: watch:tsc",
+			"dependsOn": ["npm: protos"],
 			"presentation": {
 				"group": "watch",
-				"reveal": "never"
+				"reveal": "always"
 			}
 		},
 		{
@@ -55,16 +169,30 @@
 			"script": "watch-tests",
 			"problemMatcher": "$tsc-watch",
 			"isBackground": true,
+			"dependsOn": ["npm: protos"],
 			"presentation": {
-				"reveal": "never",
+				"reveal": "always",
 				"group": "watchers"
 			},
 			"group": "build"
 		},
 		{
 			"label": "tasks: watch-tests",
-			"dependsOn": ["npm: watch", "npm: watch-tests"],
+			"dependsOn": ["npm: protos", "npm: watch", "npm: watch-tests"],
 			"problemMatcher": []
+		},
+		{
+			"label": "stop",
+			"command": "echo ${input:terminate}",
+			"type": "shell"
+		}
+	],
+	"inputs": [
+		{
+			"id": "terminate",
+			"type": "command",
+			"command": "workbench.action.tasks.terminate",
+			"args": "terminateAll"
 		}
 	]
 }

.vscodeignore

@@ -23,7 +23,6 @@ demo.gif
 # Ignore all webview-ui files except the build directory (https://github.com/microsoft/vscode-webview-ui-toolkit-samples/blob/main/frameworks/hello-world-react-cra/.vscodeignore)
 webview-ui/src/**
 webview-ui/public/**
-webview-ui/scripts/**
 webview-ui/index.html
 webview-ui/README.md
 webview-ui/package.json
@@ -38,8 +37,12 @@ docs/**
 !node_modules/@vscode/codicons/dist/codicon.css
 !node_modules/@vscode/codicons/dist/codicon.ttf
 
+# Include KaTeX CSS and fonts for LaTeX rendering
+!webview-ui/node_modules/katex/dist/katex.min.css
+!webview-ui/node_modules/katex/dist/fonts/**
+
 # Include default themes JSON files used in getTheme
 !src/integrations/theme/default-themes/**
 
 # Include icons
-!assets/icons/**
+!assets/icons/**

CHANGELOG.md

@@ -1,344 +1,637 @@
 # Changelog
 
+## [3.13.3]
+
+-   Add download counts to MCP marketplace items
+-   Add `/compact` command
+-   Add prompt caching to gemini models in cline / openrouter providers
+-   Add tooltips to bottom row menu
+
+## [3.13.2]
+
+-   Add Gemini 2.5 Flash model to Vertex and Gemini Providers (Thanks monotykamary!)
+-   Add Caching to gemini provider (Thanks arafatkatze!)
+-   Add thinking budget support to Gemini Models (Thanks monotykamary!)
+-   Add !include .file directive support for .clineignore (Thanks watany-dev!)
+-   Improve slash command functionality
+-   Improve prompting for new task tool
+-   Fix o1 temperature being passed to the azure api (Thanks treeleaves30760!)
+-   Fix to make "add new rule file" button functional
+-   Fix Ollama provider timeout, allowing for a larger loading time (Thanks suvarchal!)
+-   Fix Non-UTF-8 File Handling: Improve Encoding Detection to Prevent Garbled Text and Binary Misclassification (Thanks yt3trees!)
+-   Fix settings to not reset by changing providers
+-   Fix terminal outputs missing commas
+-   Fix terminal errors caused by starting non-alphanumeric outputs
+-   Fix auto approve settings becoming unset
+-   Fix Mermaid syntax error in documentation (Thanks tuki0918!)
+-   Remove supportsComputerUse restriction and support browser use through any model that supports images (Thanks arafatkatze!)
+
+## [3.13.1]
+
+-   Fix bug where task cancellation during thinking stream would result in error state
+
+## [3.13.0]
+
+-   Add Cline rules popover under the chat field, allowing you to easily add, enable & disable workspace level or global rule files
+-   Add new slash command menu letting you type “/“ to do quick actions like creating new tasks
+-   Add ability to edit past messages, with options to restore your workspace back to that point
+-   Allow sending a message when selecting an option provided by the question or plan tool
+-   Add command to jump to Cline's chat input
+-   Add support for OpenAI o3 & 4o-mini (Thanks @PeterDaveHello and @arafatkatze!)
+-   Add baseURL option for Google Gemini provider (Thanks @owengo and @olivierhub!)
+-   Add support for Azure's DeepSeek model. (Thanks @yt3trees!)
+-   Add ability for models that support it to receive image responses from MCP servers (Thanks @rikaaa0928!)
+-   Improve search and replace diff editing by making it more flexible with models that fail to follow structured output instructions. (Thanks @chi-cat!)
+-   Add detection of Ctrl+C termination in terminal, improving output reading issues
+-   Fix issue where some commands with large output would cause UI to freeze
+-   Fix token usage tracking issues with vertex provider (Thanks @mzsima!)
+-   Fix issue with xAI reasoning content not being parsed (Thanks @mrubens!)
+
+## [3.12.3]
+
+-   Add copy button to MermaidBlock component (Thanks @cacosub7!)
+-   Add the ability to fetch from global cline rules files
+-   Add icon to indicate when a file outside of the users workspace is edited
+
+## [3.12.2]
+
+-   Add gpt-4.1
+
+## [3.12.1]
+
+-   Use visual checkpoint indicator to make it clear when checkpoints are created
+-   Big shoutout to @samuel871211 for numerous code quality improvements, refactoring contributions, and webview performance improvements!
+-   Use improved context manager
+
+## [3.12.0]
+
+-   Add favorite toggles for models when using the Cline & OpenRouter providers
+-   Add auto-approve options for edits/reads outside of the workspace
+-   Improve diff editing animation for large files
+-   Add indicator showing number of diff edits when Cline edits a file
+-   Add streaming support and reasoning effort option to xAI's Grok 3 Mini
+-   Add settings button to MCP popover to easily modify installed servers
+-   Fix bug where browser tool actions would show unparsed results in the chat view
+-   Fix issue with new checkpoints popover hiding too quickly
+-   Fix duplicate checkpoints bug
+-   Improve Ollama provider with retry mechanism, timeout handling, and improved error handling (thanks suvarchal!)
+
+## [3.11.0]
+
+-   Redesign checkpoint UI to declutter chat view by using a subtle indicator line that expands to a popover on hover, with a new date indicator for when it was created
+-   Add support for xAI's provider's Grok 3 models
+-   Add more robust error tracking for users opted in to telemetry (thank you for helping us make Cline better!)
+
+## [3.10.1]
+
+-   Add CMD+' keyboard shortcut to add selected text to Cline
+-   Cline now auto focuses the text field when using 'Add to Cline' shortcut
+-   Add new 'Create New Task' tool to let Cline start a new task autonomously!
+-   Fix Mermaid diagram issues
+-   Fix Gemini provider cost calculation to take new tiered pricing structure into account
+
+## [3.10.0]
+
+-   Add setting to let browser tool use local Chrome via remote debugging, enabling session-based browsing. Replaces sessionless Chromium, unlocking debugging and productivity workflows tied to your real browser state.
+-   Add new auto-approve option to approve _ALL_ commands (use at your own risk!)
+-   Add modal in the chat area to more easily enable or disable MCP servers
+-   Add drag and drop of file/folders into cline chat (Thanks eljapi!)
+-   Add prompt caching for LiteLLM + Claude (Thanks sammcj!)
+-   Add Improved context management
+-   Fix MCP auto approve toggle issues being out of sync with settings
+
+## [3.9.2]
+
+-   Add recommended models for Cline provider
+-   Add ability to detect when user edits files manually so Cline knows to re-read, leading to reduced diff edit errors
+-   Add improvements to file mention searching for faster searching
+-   Add scoring logic to file mentions to sort and exclude results based on relevance
+-   Add Support for Bytedance Doubao (Thanks Tunixer!)
+-   Fix to prevent duplicate BOM (Thanks bamps53!)
+
+## [3.9.1]
+
+-   Add Gemini 2.5 Pro Preview 03-25 to Google Provider
+
+## [3.9.0]
+
+-   Add Enable extended thinking for LiteLLM provider (Thanks @jorgegarciarey!)
+-   Add a tab for configuring local MCP Servers
+-   Fix issue with DeepSeek API provider token counting + context management
+-   Fix issues with checkpoints hanging under certain conditions
+
+## [3.8.6]
+
+-   Add UI for adding remote servers
+-   Add Mentions Feature Guide and update related documentation
+-   Fix bug where menu would open in sidebar and open tab
+-   Fix issue with Cline accounts not showing user info in popout tabs
+-   Fix bug where menu buttons wouldn't open view in sidebar
+
+## [3.8.5]
+
+-   Add support for remote MCP Servers using SSE
+-   Add gemini-2.5-pro-exp-03-25 to Vertex AI (thanks @arri-cc!)
+-   Add access to history, mcp, and new task buttons in popout view
+-   Add task feedback telemetry (thumbs up/down on task completion)
+-   Add toggle disabled for remote servers
+-   Move the MCP Restart and Delete buttons and add an auto-approve all toggle
+-   Update Requestly UX for model selection (thanks @arafatkatze!)
+-   Add escape for html content for gemini when running commands
+-   Improve search and replace edit failure behaviors
+
+## [3.8.4]
+
+-   Add Sambanova Deepseek-V3-0324
+-   Add cost calculation support for LiteLLM provider
+-   Fix bug where Cline would use plan_mode_response bug without response parameter
+
+## [3.8.3]
+
+-   Add support for SambaNova QwQ-32B model
+-   Add OpenAI "dynamic" model chatgpt-4o-latest
+-   Add Amazon Nova models to AWS Bedrock
+-   Improve file handling for NextJS folder naming (fixes issues with parentheses in folder names)
+-   Add Gemini 2.5 Pro to Google AI Studio available models
+-   Handle "input too large" errors for Anthropic
+-   Fix "See more" not showing up for tasks after task un-fold
+-   Fix gpt-4.5-preview's supportsPromptCache value to true
+
+## [3.8.2]
+
+-   Fix bug where switching to plan/act would result in VS Code LM/OpenRouter model being reset
+
+## [3.8.0]
+
+-   Add 'Add to Cline' as an option when you right-click in a file or the terminal, making it easier to add context to your current task
+-   Add 'Fix with Cline' code action - when you see a lightbulb icon in your editor, you can now select 'Fix with Cline' to send the code and associated errors for Cline to fix. (Cursor users can also use the 'Quick Fix (CMD + .)' menu to see this option)
+-   Add Account view to display billing and usage history for Cline account users. You can now keep track of credits used and transaction history right in the extension!
+-   Add 'Sort underling provider routing' setting to Cline/OpenRouter allowing you to sort provider used by throughput, price, latency, or the default (combination of price and uptime)
+-   Improve rich MCP display with dynamic image loading and support for GIFs
+-   Add 'Documentation' menu item to easily access Cline's docs
+-   Add OpenRouter's new usage_details feature for more reliable cost reporting
+-   Display total space Cline takes on disk next to 'Delete all Tasks' button in History view
+-   Fix 'Context Window Exceeded' error for OpenRouter/Cline Accounts (additional support coming soon)
+-   Fix bug where OpenRouter model ID would be set to invalid value
+-   Add button to delete MCP servers in a failure state
+
+## [3.7.1]
+
+-   Fix issue with 'See more' button in task header not showing when starting new tasks
+-   Fix issue with checkpoints using local git commit hooks
+
+## [3.7.0]
+
+-   Cline now displays selectable options when asking questions or presenting a plan, saving you from having to type out responses!
+-   Add support for a `.clinerules/` directory to load multiple files at once (thanks @ryo-ma!)
+-   Prevent Cline from reading extremely large files into context that would overload context window
+-   Improve checkpoints loading performance and display warning for large projects not suited for checkpoints
+-   Add SambaNova API provider (thanks @saad-noodleseed!)
+-   Add VPC endpoint option for AWS Bedrock profiles (thanks @minorunara!)
+-   Add DeepSeek-R1 to AWS Bedrock (thanks @watany-dev!)
+
+## [3.6.5]
+
+-   Add 'Delete all Task History' button to History view
+-   Add toggle to disable model switching between Plan/Act modes in Settings (new users default to disabled)
+-   Add temperature option to OpenAI Compatible
+-   Add Kotlin support to tree-sitter parser (thanks @fumiya-kume!)
+
+## [3.6.3]
+
+-   Improve QwQ support for Alibaba (thanks @meglinge!) and OpenRouter
+-   Improve diff edit prompting to prevent immediately reverting to write_to_file when a model uses search patterns that don't match anything in the file
+-   Fix bug where new checkpoints system would revert file changes when switching between tasks
+-   Fix issue with incorrect token count for some OpenAI compatible providers
+
+## [3.6.0]
+
+-   Add Cline API as a provider option, allowing new users to sign up and get started with Cline for free
+-   Optimize checkpoints with branch-per-task strategy, reducing storage required and first task load times
+-   Fix problem with Plan/Act toggle keyboard shortcut not working in Windows (thanks @yt3trees!)
+-   Add new Gemini models to GCP Vertex (thanks @shohei-ihaya!) and Claude models AskSage (thanks @swhite24!)
+-   Improve OpenRouter/Cline error reporting
+
+## [3.5.1]
+
+-   Add timeout option to MCP servers
+-   Add Gemini Flash models to Vertex provider (thanks @jpaodev!)
+-   Add prompt caching support for AWS Bedrock provider (thanks @buger!)
+-   Add AskSage provider (thanks @swhite24!)
+
+## [3.5.0]
+
+-   Add 'Enable extended thinking' option for Claude 3.7 Sonnet, with ability to set different budgets for Plan and Act modes
+-   Add support for rich MCP responses with automatic image previews, website thumbnails, and WolframAlpha visualizations
+-   Add language preference option in Advanced Settings
+-   Add xAI Provider Integration with support for all Grok models (thanks @andrewmonostate!)
+-   Fix issue with Linux XDG pointing to incorrect path for Document folder (thanks @jonatkinson!)
+
+## [3.4.10]
+
+-   Add support for GPT-4.5 preview model
+
+## [3.4.9]
+
+-   Add toggle to let users opt-in to anonymous telemetry and error reporting
+
+## [3.4.6]
+
+-   Add support for Claude 3.7 Sonnet
+
+## [3.4.0]
+
+-   Introducing MCP Marketplace! You can now discover and install the best MCP servers right from within the extension, with new servers added regularly
+-   Add mermaid diagram support in Plan mode! You can now see visual representations of mermaid code blocks in chat, and click on them to see an expanded view
+-   Use more visual checkpoints indicators after editing files & running commands
+-   Create a checkpoint at the beginning of each task to easily revert to the initial state
+-   Add 'Terminal' context mention to reference the active terminal's contents
+-   Add 'Git Commits' context mention to reference current working changes or specific commits (thanks @mrubens!)
+-   Send current textfield contents as additional feedback when toggling from Plan to Act Mode, or when hitting 'Approve' button
+-   Add advanced configuration options for OpenAI Compatible (context window, max output, pricing, etc.)
+-   Add Alibaba Qwen 2.5 coder models, VL models, and DeepSeek-R1/V3 support
+-   Improve support for AWS Bedrock Profiles
+-   Fix Mistral provider support for non-codestral models
+-   Add advanced setting to disable browser tool
+-   Add advanced setting to set chromium executable path for browser tool
+
+## [3.3.2]
+
+-   Fix bug where OpenRouter requests would periodically not return cost/token stats, leading to context window limit errors
+-   Make checkpoints more visible and keep track of restored checkpoints
+
+## [3.3.0]
+
+-   Add .clineignore to block Cline from accessing specified file patterns
+-   Add keyboard shortcut + tooltips for Plan/Act toggle
+-   Fix bug where new files won't show up in files dropdown
+-   Add automatic retry for rate limited requests (thanks @ViezeVingertjes!)
+-   Adding reasoning_effort support for o3-mini in Advanced Settings
+-   Added support for AWS provider profiles using the AWS CLI to make the profile, enabling long lived connections to AWS bedrock
+-   Adding Requesty API provider
+-   Add Together API provider
+-   Add Alibaba Qwen API provider (thanks @aicccode!)
+
+## [3.2.13]
+
+-   Add new gemini models gemini-2.0-flash-lite-preview-02-05 and gemini-2.0-flash-001
+-   Add all available Mistral API models (thanks @ViezeVingertjes!)
+-   Add LiteLLM API provider support (thanks @him0!)
+
+## [3.2.12]
+
+-   Fix command chaining for Windows users
+-   Fix reasoning_content error for OpenAI providers
+
+## [3.2.11]
+
+-   Add OpenAI o3-mini model
+
+## [3.2.10]
+
+-   Improve support for DeepSeek-R1 (deepseek-reasoner) model for OpenRouter, OpenAI-compatible, and DeepSeek direct (thanks @Szpadel!)
+-   Show Reasoning tokens for models that support it
+-   Fix issues with switching models between Plan/Act modes
+
 ## [3.2.6]
 
-- Save last used API/model when switching between Plan and Act, for users that like to use different models for each mode
-- New Context Window progress bar in the task header to understand increased cost/generation degradation as the context increases
-- Localize READMEs and add language selector for English, Spanish, German, Chinese, and Japanese
-- Add Advanced Settings to remove MCP prompts from requests to save tokens, enable/disable checkpoints for users that don't use git (more coming soon!)
-- Add Gemini 2.0 Flash Thinking experimental model
-- Allow new users to subscribe to mailing list to get notified when new Accounts option is available
+-   Save last used API/model when switching between Plan and Act, for users that like to use different models for each mode
+-   New Context Window progress bar in the task header to understand increased cost/generation degradation as the context increases
+-   Localize READMEs and add language selector for English, Spanish, German, Chinese, and Japanese
+-   Add Advanced Settings to remove MCP prompts from requests to save tokens, enable/disable checkpoints for users that don't use git (more coming soon!)
+-   Add Gemini 2.0 Flash Thinking experimental model
+-   Allow new users to subscribe to mailing list to get notified when new Accounts option is available
 
 ## [3.2.5]
 
-- Use yellow textfield outline in Plan mode to better distinguish from Act mode
+-   Use yellow textfield outline in Plan mode to better distinguish from Act mode
 
 ## [3.2.3]
 
-- Add DeepSeek-R1 (deepseek-reasoner) model support with proper parameter handling (thanks @slavakurilyak!)
+-   Add DeepSeek-R1 (deepseek-reasoner) model support with proper parameter handling (thanks @slavakurilyak!)
 
 ## [3.2.0]
 
-- Add Plan/Act mode toggle to let you plan tasks with Cline before letting him get to work
-- Easily switch between API providers and models using a new popup menu under the chat field
-- Add VS Code LM API provider to run models provided by other VS Code extensions (e.g. GitHub Copilot). Shoutout to @julesmons, @RaySinner, and @MrUbens for putting this together!
-- Add on/off toggle for MCP servers to disable them when not in use. Thanks @MrUbens!
-- Add Auto-approve option for individual tools in MCP servers. Thanks @MrUbens!
+-   Add Plan/Act mode toggle to let you plan tasks with Cline before letting him get to work
+-   Easily switch between API providers and models using a new popup menu under the chat field
+-   Add VS Code LM API provider to run models provided by other VS Code extensions (e.g. GitHub Copilot). Shoutout to @julesmons, @RaySinner, and @MrUbens for putting this together!
+-   Add on/off toggle for MCP servers to disable them when not in use. Thanks @MrUbens!
+-   Add Auto-approve option for individual tools in MCP servers. Thanks @MrUbens!
 
 ## [3.1.10]
 
-- New icon!
+-   New icon!
 
 ## [3.1.9]
 
-- Add Mistral API provider with codestral-latest model
+-   Add Mistral API provider with codestral-latest model
 
 ## [3.1.7]
 
-- Add ability to change viewport size and headless mode when Cline asks to launch the browser
+-   Add ability to change viewport size and headless mode when Cline asks to launch the browser
 
 ## [3.1.6]
 
-- Fix bug where filepaths with Chinese characters would not show up in context mention menu (thanks @chi-chat!)
-- Update Anthropic model prices (thanks @timoteostewart!)
+-   Fix bug where filepaths with Chinese characters would not show up in context mention menu (thanks @chi-chat!)
+-   Update Anthropic model prices (thanks @timoteostewart!)
 
 ## [3.1.5]
 
-- Fix bug where Cline couldn't read "@/" import path aliases from tool results
+-   Fix bug where Cline couldn't read "@/" import path aliases from tool results
 
 ## [3.1.4]
 
-- Fix issue where checkpoints would not work for users with git commit signing enabled globally
+-   Fix issue where checkpoints would not work for users with git commit signing enabled globally
 
 ## [3.1.2]
 
-- Fix issue where LFS files would be not be ignored when creating checkpoints
+-   Fix issue where LFS files would be not be ignored when creating checkpoints
 
 ## [3.1.0]
 
-- Added checkpoints: Snapshots of workspace are automatically created whenever Cline uses a tool
-  - Compare changes: Hover over any tool use to see a diff between the snapshot and current workspace state
-  - Restore options: Choose to restore just the task state, just the workspace files, or both
-- New 'See new changes' button appears after task completion, providing an overview of all workspace changes
-- Task header now shows disk space usage with a delete button to help manage snapshot storage
+-   Added checkpoints: Snapshots of workspace are automatically created whenever Cline uses a tool
+-   Compare changes: Hover over any tool use to see a diff between the snapshot and current workspace state
+-   Restore options: Choose to restore just the task state, just the workspace files, or both
+-   New 'See new changes' button appears after task completion, providing an overview of all workspace changes
+-   Task header now shows disk space usage with a delete button to help manage snapshot storage
 
 ## [3.0.12]
 
-- Fix DeepSeek API cost reporting (input price is 0 since it's all either a cache read or write, different than how Anthropic reports cache usage)
+-   Fix DeepSeek API cost reporting (input price is 0 since it's all either a cache read or write, different than how Anthropic reports cache usage)
 
 ## [3.0.11]
 
-- Emphasize auto-formatting done by the editor in file edit responses for more reliable diff editing
+-   Emphasize auto-formatting done by the editor in file edit responses for more reliable diff editing
 
 ## [3.0.10]
 
-- Add DeepSeek provider to API Provider options
-- Fix context window limit errors for DeepSeek v3
+-   Add DeepSeek provider to API Provider options
+-   Fix context window limit errors for DeepSeek v3
 
 ## [3.0.9]
 
-- Fix bug where DeepSeek v3 would incorrectly escape HTML entities in diff edits
+-   Fix bug where DeepSeek v3 would incorrectly escape HTML entities in diff edits
 
 ## [3.0.8]
 
-- Mitigate DeepSeek v3 diff edit errors by adding 'auto-formatting considerations' to system prompt, encouraging model to use updated file contents as reference point for SEARCH blocks
+-   Mitigate DeepSeek v3 diff edit errors by adding 'auto-formatting considerations' to system prompt, encouraging model to use updated file contents as reference point for SEARCH blocks
 
 ## [3.0.7]
 
-- Revert to using batched file watcher to fix crash when many files would be created at once
+-   Revert to using batched file watcher to fix crash when many files would be created at once
 
 ## [3.0.6]
 
-- Fix bug where some files would be missing in the `@` context mention menu
-- Add Bedrock support in additional regions
-- Diff edit improvements
-- Add OpenRouter's middle-out transform for models that don't use prompt caching (prevents context window limit errors, but cannot be applied to models like Claude since it would continuously break the cache)
+-   Fix bug where some files would be missing in the `@` context mention menu
+-   Add Bedrock support in additional regions
+-   Diff edit improvements
+-   Add OpenRouter's middle-out transform for models that don't use prompt caching (prevents context window limit errors, but cannot be applied to models like Claude since it would continuously break the cache)
 
 ## [3.0.4]
 
-- Fix bug where gemini models would add code block artifacts to the end of text content
-- Fix context mention menu visual issues on light themes
+-   Fix bug where gemini models would add code block artifacts to the end of text content
+-   Fix context mention menu visual issues on light themes
 
 ## [3.0.2]
 
-- Adds block anchor matching for more reliable diff edits (if 3+ lines, first and last line are used as anchors to search for)
-- Add instruction to system prompt to use complete lines in diff edits to work properly with fallback strategies
-- Improves diff edit error handling
-- Adds new Gemini models
+-   Adds block anchor matching for more reliable diff edits (if 3+ lines, first and last line are used as anchors to search for)
+-   Add instruction to system prompt to use complete lines in diff edits to work properly with fallback strategies
+-   Improves diff edit error handling
+-   Adds new Gemini models
 
 ## [3.0.0]
 
-- Cline now uses a search & replace diff based approach when editing large files to prevent code deletion issues.
-- Adds support for a more comprehensive auto-approve configuration, allowing you to specify which tools require approval and which don't.
-- Adds ability to enable system notifications for when Cline needs approval or completes a task.
-- Adds support for a root-level `.clinerules` file that can be used to specify custom instructions for the project.
+-   Cline now uses a search & replace diff based approach when editing large files to prevent code deletion issues.
+-   Adds support for a more comprehensive auto-approve configuration, allowing you to specify which tools require approval and which don't.
+-   Adds ability to enable system notifications for when Cline needs approval or completes a task.
+-   Adds support for a root-level `.clinerules` file that can be used to specify custom instructions for the project.
 
 ## [2.2.0]
 
-- Add support for Model Context Protocol (MCP), enabling Cline to use custom tools like web-search tool or GitHub tool
-- Add MCP server management tab accessible via the server icon in the menu bar
-- Add ability for Cline to dynamically create new MCP servers based on user requests (e.g., "add a tool that gets the latest npm docs")
+-   Add support for Model Context Protocol (MCP), enabling Cline to use custom tools like web-search tool or GitHub tool
+-   Add MCP server management tab accessible via the server icon in the menu bar
+-   Add ability for Cline to dynamically create new MCP servers based on user requests (e.g., "add a tool that gets the latest npm docs")
 
 ## [2.1.6]
 
-- Add LM Studio as an API provider option (make sure to start the LM Studio server to use it with the extension!)
+-   Add LM Studio as an API provider option (make sure to start the LM Studio server to use it with the extension!)
 
 ## [2.1.5]
 
-- Add support for prompt caching for new Claude model IDs on OpenRouter (e.g. `anthropic/claude-3.5-sonnet-20240620`)
+-   Add support for prompt caching for new Claude model IDs on OpenRouter (e.g. `anthropic/claude-3.5-sonnet-20240620`)
 
 ## [2.1.4]
 
-- AWS Bedrock fixes (add missing regions, support for cross-region inference, and older Sonnet model for regions where new model is not available)
+-   AWS Bedrock fixes (add missing regions, support for cross-region inference, and older Sonnet model for regions where new model is not available)
 
 ## [2.1.3]
 
-- Add support for Claude 3.5 Haiku, 66% cheaper than Sonnet with similar intelligence
+-   Add support for Claude 3.5 Haiku, 66% cheaper than Sonnet with similar intelligence
 
 ## [2.1.2]
 
-- Misc. bug fixes
-- Update README with new browser feature
+-   Misc. bug fixes
+-   Update README with new browser feature
 
 ## [2.1.1]
 
-- Add stricter prompt to prevent Cline from editing files during a browser session without first closing the browser
+-   Add stricter prompt to prevent Cline from editing files during a browser session without first closing the browser
 
 ## [2.1.0]
 
-- Cline now uses Anthropic's new "Computer Use" feature to launch a browser, click, type, and scroll. This gives him more autonomy in runtime debugging, end-to-end testing, and even general web use. Try asking "Look up the weather in Colorado" to see it in action! (Available with Claude 3.5 Sonnet v2)
+-   Cline now uses Anthropic's new "Computer Use" feature to launch a browser, click, type, and scroll. This gives him more autonomy in runtime debugging, end-to-end testing, and even general web use. Try asking "Look up the weather in Colorado" to see it in action! (Available with Claude 3.5 Sonnet v2)
 
 ## [2.0.19]
 
-- Fix model info for Claude 3.5 Sonnet v1 on OpenRouter
+-   Fix model info for Claude 3.5 Sonnet v1 on OpenRouter
 
 ## [2.0.18]
 
-- Add support for both v1 and v2 of Claude 3.5 Sonnet for GCP Vertex and AWS Bedrock (for cases where the new model is not enabled yet or unavailable in your region)
+-   Add support for both v1 and v2 of Claude 3.5 Sonnet for GCP Vertex and AWS Bedrock (for cases where the new model is not enabled yet or unavailable in your region)
 
 ## [2.0.17]
 
-- Update Anthropic model IDs
+-   Update Anthropic model IDs
 
 ## [2.0.16]
 
-- Adjustments to system prompt
+-   Adjustments to system prompt
 
 ## [2.0.15]
 
-- Fix bug where modifying Cline's edits would lead him to try to re-apply the edits
-- Fix bug where weaker models would display file contents before using the write_to_file tool
-- Fix o1-mini and o1-preview errors when using OpenAI native
+-   Fix bug where modifying Cline's edits would lead him to try to re-apply the edits
+-   Fix bug where weaker models would display file contents before using the write_to_file tool
+-   Fix o1-mini and o1-preview errors when using OpenAI native
 
 ## [2.0.14]
 
-- Gracefully cancel requests while stream could be hanging
+-   Gracefully cancel requests while stream could be hanging
 
 ## [2.0.13]
 
-- Detect code omission and show warning with troubleshooting link
+-   Detect code omission and show warning with troubleshooting link
 
 ## [2.0.12]
 
-- Keep cursor out of the way during file edit streaming animation
+-   Keep cursor out of the way during file edit streaming animation
 
 ## [2.0.11]
 
-- Adjust prompts around read_file to prevent re-reading files unnecessarily
+-   Adjust prompts around read_file to prevent re-reading files unnecessarily
 
 ## [2.0.10]
 
-- More adjustments to system prompt to prevent lazy coding
+-   More adjustments to system prompt to prevent lazy coding
 
 ## [2.0.9]
 
-- Update system prompt to try to prevent Cline from lazy coding (`// rest of code here...`)
+-   Update system prompt to try to prevent Cline from lazy coding (`// rest of code here...`)
 
 ## [2.0.8]
 
-- Fix o1-mini and o1-preview for OpenAI
-- Fix diff editor not opening sometimes in slow environments like project idx
+-   Fix o1-mini and o1-preview for OpenAI
+-   Fix diff editor not opening sometimes in slow environments like project idx
 
 ## [2.0.7]
 
-- Misc. bug fixes
+-   Misc. bug fixes
 
 ## [2.0.6]
 
-- Update URLs to https://github.com/cline/cline
+-   Update URLs to https://github.com/cline/cline
 
 ## [2.0.5]
 
-- Fixed bug where Cline's edits would stream into the active tab when switching tabs during a write_to_file
-- Added explanation in task continuation prompt that an interrupted write_to_file reverts the file to its original contents, preventing unnecessary re-reads
-- Fixed non-first chunk error handling in case stream fails mid-way through
+-   Fixed bug where Cline's edits would stream into the active tab when switching tabs during a write_to_file
+-   Added explanation in task continuation prompt that an interrupted write_to_file reverts the file to its original contents, preventing unnecessary re-reads
+-   Fixed non-first chunk error handling in case stream fails mid-way through
 
 ## [2.0.0]
 
-- New name! Meet Cline, an AI assistant that can use your CLI and Editor
-- Responses are now streamed with a yellow text decoration animation to keep track of Cline's progress as he edits files
-- New Cancel button to give Cline feedback if he goes off in the wrong direction, giving you more control over tasks
-- Re-imagined tool calling prompt resulting in ~40% fewer requests to accomplish tasks + better performance with other models
-- Search and use any model with OpenRouter
+-   New name! Meet Cline, an AI assistant that can use your CLI and Editor
+-   Responses are now streamed with a yellow text decoration animation to keep track of Cline's progress as he edits files
+-   New Cancel button to give Cline feedback if he goes off in the wrong direction, giving you more control over tasks
+-   Re-imagined tool calling prompt resulting in ~40% fewer requests to accomplish tasks + better performance with other models
+-   Search and use any model with OpenRouter
 
 ## [1.9.7]
 
-- Only auto-include error diagnostics after file edits, removed warnings to keep Claude from getting distracted in projects with strict linting rules
+-   Only auto-include error diagnostics after file edits, removed warnings to keep Claude from getting distracted in projects with strict linting rules
 
 ## [1.9.6]
 
-- Added support for new Google Gemini models `gemini-1.5-flash-002` and `gemini-1.5-pro-002`
-- Updated system prompt to be more lenient when terminal output doesn't stream back properly
-- Adjusted system prompt to prevent overuse of the inspect_site tool
-- Increased global line height for improved readability
+-   Added support for new Google Gemini models `gemini-1.5-flash-002` and `gemini-1.5-pro-002`
+-   Updated system prompt to be more lenient when terminal output doesn't stream back properly
+-   Adjusted system prompt to prevent overuse of the inspect_site tool
+-   Increased global line height for improved readability
 
 ## [1.9.0]
 
-- Claude can now use a browser! This update adds a new `inspect_site` tool that captures screenshots and console logs from websites (including localhost), making it easier for Claude to troubleshoot issues on his own.
-- Improved automatic linter/compiler debugging by only sending Claude new errors that result from his edits, rather than reporting all workspace problems.
+-   Claude can now use a browser! This update adds a new `inspect_site` tool that captures screenshots and console logs from websites (including localhost), making it easier for Claude to troubleshoot issues on his own.
+-   Improved automatic linter/compiler debugging by only sending Claude new errors that result from his edits, rather than reporting all workspace problems.
 
 ## [1.8.0]
 
-- You can now use '@' in the textarea to add context!
-    - @url: Paste in a URL for the extension to fetch and convert to markdown, useful when you want to give Claude the latest docs!
-    - @problems: Add workspace errors and warnings for Claude to fix, no more back-and-forth about debugging
-    - @file: Adds a file's contents so you don't have to waste API requests approving read file (+ type to search files)
-    - @folder: Adds folder's files all at once to speed up your workflow even more
+-   You can now use '@' in the textarea to add context!
+-   @url: Paste in a URL for the extension to fetch and convert to markdown, useful when you want to give Claude the latest docs!
+-   @problems: Add workspace errors and warnings for Claude to fix, no more back-and-forth about debugging
+-   @file: Adds a file's contents so you don't have to waste API requests approving read file (+ type to search files)
+-   @folder: Adds folder's files all at once to speed up your workflow even more
 
 ## [1.7.0]
 
-- Adds problems monitoring to keep Claude updated on linter/compiler/build issues, letting him proactively fix errors on his own! (adding missing imports, fixing type errors, etc.)
+-   Adds problems monitoring to keep Claude updated on linter/compiler/build issues, letting him proactively fix errors on his own! (adding missing imports, fixing type errors, etc.)
 
 ## [1.6.5]
 
-- Adds support for OpenAI o1, Azure OpenAI, and Google Gemini (free for up to 15 requests per minute!)
-- Task header can now be collapsed to provide more space for viewing conversations
-- Adds fuzzy search and sorting to Task History, making it easier to find specific tasks
+-   Adds support for OpenAI o1, Azure OpenAI, and Google Gemini (free for up to 15 requests per minute!)
+-   Task header can now be collapsed to provide more space for viewing conversations
+-   Adds fuzzy search and sorting to Task History, making it easier to find specific tasks
 
 ## [1.6.0]
 
-- Commands now run directly in your terminal thanks to VSCode 1.93's new shell integration updates! Plus a new 'Proceed While Running' button to let Claude continue working while commands run, sending him new output along the way (i.e. letting him react to server errors as he edits files)
+-   Commands now run directly in your terminal thanks to VSCode 1.93's new shell integration updates! Plus a new 'Proceed While Running' button to let Claude continue working while commands run, sending him new output along the way (i.e. letting him react to server errors as he edits files)
 
 ## [1.5.27]
 
-- Claude's changes now appear in your file's Timeline, allowing you to easily view a diff of each edit. This is especially helpful if you want to revert to a previous version. No need for git—everything is tracked by VSCode's local history!
-- Updated system prompt to keep Claude from re-reading files unnecessarily
+-   Claude's changes now appear in your file's Timeline, allowing you to easily view a diff of each edit. This is especially helpful if you want to revert to a previous version. No need for git—everything is tracked by VSCode's local history!
+-   Updated system prompt to keep Claude from re-reading files unnecessarily
 
 ## [1.5.19]
 
-- Adds support for OpenAI compatible API providers (e.g. Ollama!)
+-   Adds support for OpenAI compatible API providers (e.g. Ollama!)
 
 ## [1.5.13]
 
-- New terminal emulator! When Claude runs commands, you can now type directly in the terminal (+ support for Python environments)
-- Adds search to Task History
+-   New terminal emulator! When Claude runs commands, you can now type directly in the terminal (+ support for Python environments)
+-   Adds search to Task History
 
 ## [1.5.6]
 
-- You can now edit Claude's changes before accepting! When he edits or creates a file, you can modify his changes directly in the right side of the diff view (+ hover over the 'Revert Block' arrow button in the center to undo `// rest of code here` shenanigans)
+-   You can now edit Claude's changes before accepting! When he edits or creates a file, you can modify his changes directly in the right side of the diff view (+ hover over the 'Revert Block' arrow button in the center to undo `// rest of code here` shenanigans)
 
 ## [1.5.4]
 
-- Adds support for reading .pdf and .docx files (try "turn my business_plan.docx into a company website")
+-   Adds support for reading .pdf and .docx files (try "turn my business_plan.docx into a company website")
 
 ## [1.5.0]
 
-- Adds new `search_files` tool that lets Claude perform regex searches in your project, making it easy for him to refactor code, address TODOs and FIXMEs, remove dead code, and more!
+-   Adds new `search_files` tool that lets Claude perform regex searches in your project, making it easy for him to refactor code, address TODOs and FIXMEs, remove dead code, and more!
 
 ## [1.4.0]
 
-- Adds "Always allow read-only operations" setting to let Claude read files and view directories without needing approval (off by default)
-- Implement sliding window context management to keep tasks going past 200k tokens
-- Adds Google Cloud Vertex AI support and updates Claude 3.5 Sonnet max output to 8192 tokens for all providers.
-- Improves system prompt to gaurd against lazy edits (less "//rest of code here")
+-   Adds "Always allow read-only operations" setting to let Claude read files and view directories without needing approval (off by default)
+-   Implement sliding window context management to keep tasks going past 200k tokens
+-   Adds Google Cloud Vertex AI support and updates Claude 3.5 Sonnet max output to 8192 tokens for all providers.
+-   Improves system prompt to guard against lazy edits (less "//rest of code here")
 
 ## [1.3.0]
 
-- Adds task history
+-   Adds task history
 
 ## [1.2.0]
 
-- Adds support for Prompt Caching to significantly reduce costs and response times (currently only available through Anthropic API for Claude 3.5 Sonnet and Claude 3.0 Haiku)
+-   Adds support for Prompt Caching to significantly reduce costs and response times (currently only available through Anthropic API for Claude 3.5 Sonnet and Claude 3.0 Haiku)
 
 ## [1.1.1]
 
-- Adds option to choose other Claude models (+ GPT-4o, DeepSeek, and Mistral if you use OpenRouter)
-- Adds option to add custom instructions to the end of the system prompt
+-   Adds option to choose other Claude models (+ GPT-4o, DeepSeek, and Mistral if you use OpenRouter)
+-   Adds option to add custom instructions to the end of the system prompt
 
 ## [1.1.0]
 
-- Paste images in chat to use Claude's vision capabilities and turn mockups into fully functional applications or fix bugs with screenshots
+-   Paste images in chat to use Claude's vision capabilities and turn mockups into fully functional applications or fix bugs with screenshots
 
 ## [1.0.9]
 
-- Add support for OpenRouter and AWS Bedrock
+-   Add support for OpenRouter and AWS Bedrock
 
 ## [1.0.8]
 
-- Shows diff view of new or edited files right in the editor
+-   Shows diff view of new or edited files right in the editor
 
 ## [1.0.7]
 
-- Replace `list_files` and `analyze_project` with more explicit `list_files_top_level`, `list_files_recursive`, and `view_source_code_definitions_top_level` to get source code definitions only for files relevant to the task
+-   Replace `list_files` and `analyze_project` with more explicit `list_files_top_level`, `list_files_recursive`, and `view_source_code_definitions_top_level` to get source code definitions only for files relevant to the task
 
 ## [1.0.6]
 
-- Interact with CLI commands by sending messages to stdin and terminating long-running processes like servers
-- Export tasks to markdown files (useful as context for future tasks)
+-   Interact with CLI commands by sending messages to stdin and terminating long-running processes like servers
+-   Export tasks to markdown files (useful as context for future tasks)
 
 ## [1.0.5]
 
-- Claude now has context about vscode's visible editors and opened tabs
+-   Claude now has context about vscode's visible editors and opened tabs
 
 ## [1.0.4]
 
-- Open in the editor (using menu bar or `Claude Dev: Open In New Tab` in command palette) to see how Claude updates your workspace more clearly
-- New `analyze_project` tool to help Claude get a comprehensive overview of your project's source code definitions and file structure
-- Provide feedback to tool use like terminal commands and file edits
-- Updated max output tokens to 8192 so less lazy coding (`// rest of code here...`)
-- Added ability to retry failed API requests (helpful for rate limits)
-- Quality of life improvements like markdown rendering, memory optimizations, better theme support
+-   Open in the editor (using menu bar or `Claude Dev: Open In New Tab` in command palette) to see how Claude updates your workspace more clearly
+-   New `analyze_project` tool to help Claude get a comprehensive overview of your project's source code definitions and file structure
+-   Provide feedback to tool use like terminal commands and file edits
+-   Updated max output tokens to 8192 so less lazy coding (`// rest of code here...`)
+-   Added ability to retry failed API requests (helpful for rate limits)
+-   Quality of life improvements like markdown rendering, memory optimizations, better theme support
 
 ## [0.0.6]
 
-- Initial release
+-   Initial release

CONTRIBUTING.md

@@ -31,6 +31,36 @@ If you're planning to work on a bigger feature, please create a [feature request
     - Run `npm run test` to run tests locally
     - Before submitting PR, run `npm run format:fix` to format your code
 
+3. **Linux-specific Setup**
+    VS Code extension tests on Linux require the following system libraries:
+
+    - `libatk1.0-0`
+    - `libatk-bridge2.0-0`
+    - `libxkbfile1`
+    - `libx11-xcb1`
+    - `libxcomposite1`
+    - `libxdamage1`
+    - `libxfixes3`
+    - `libxrandr2`
+    - `libgbm1`
+    - `libdrm2`
+    - `libgtk-3-0`
+    - `dbus`
+    - `xvfb`
+
+    These libraries provide necessary GUI components and system services for the test environment.
+
+    For example, on Debian-based distributions (e.g., Ubuntu), you can install these libraries using apt:
+    ```bash
+    sudo apt update
+    sudo apt install -y \
+      libatk1.0-0 libatk-bridge2.0-0 libxkbfile1 libx11-xcb1 \
+      libxcomposite1 libxdamage1 libxfixes3 libxrandr2 libgbm1 \
+      libdrm2 libgtk-3-0 dbus xvfb
+    ```
+
+    - Run `npm run test:ci` to run tests locally
+
 ## Writing and Submitting Code
 
 Anyone can contribute code to Cline, but we ask that you follow these guidelines to ensure your contributions can be smoothly integrated:
@@ -56,20 +86,30 @@ Anyone can contribute code to Cline, but we ask that you follow these guidelines
     - Update existing tests if your changes affect them
     - Include both unit tests and integration tests where appropriate
 
-4. **Commit Guidelines**
+4. **Version Management with Changesets**
+
+    - Create a changeset for any user-facing changes using `npm run changeset`
+    - Choose the appropriate version bump:
+        - `major` for breaking changes (1.0.0 → 2.0.0)
+        - `minor` for new features (1.0.0 → 1.1.0)
+        - `patch` for bug fixes (1.0.0 → 1.0.1)
+    - Write clear, descriptive changeset messages that explain the impact
+    - Documentation-only changes don't require changesets
+
+5. **Commit Guidelines**
 
     - Write clear, descriptive commit messages
     - Use conventional commit format (e.g., "feat:", "fix:", "docs:")
     - Reference relevant issues in commits using #issue-number
 
-5. **Before Submitting**
+6. **Before Submitting**
 
     - Rebase your branch on the latest main
     - Ensure your branch builds successfully
     - Double-check all tests are passing
     - Review your changes for any debugging code or console logs
 
-6. **Pull Request Description**
+7. **Pull Request Description**
     - Clearly describe what your changes do
     - Include steps to test the changes
     - List any breaking changes

LICENSE

@@ -186,7 +186,7 @@
       same "printed page" as the copyright notice for easier
       identification within third-party archives.
 
-   Copyright 2024 Cline Bot Inc.
+   Copyright 2025 Cline Bot Inc.
 
    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.
@@ -198,4 +198,4 @@
    distributed under the License is distributed on an "AS IS" BASIS,
    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    See the License for the specific language governing permissions and
-   limitations under the License.
+   limitations under the License.

README.md

@@ -1,5 +1,5 @@
 <div align="center"><sub>
-English | <a href="https://github.com/cline/cline/blob/main/locales/es/README.md" target="_blank">Español</a> | <a href="https://github.com/cline/cline/blob/main/locales/de/README.md" target="_blank">Deutsch</a> | <a href="https://github.com/cline/cline/blob/main/locales/ja/README.md" target="_blank">日本語</a> | <a href="https://github.com/cline/cline/blob/main/locales/zh-cn/README.md" target="_blank">简体中文</a> | <a href="https://github.com/cline/cline/blob/main/locales/zh-tw/README.md" target="_blank">繁體中文</a>
+English | <a href="https://github.com/cline/cline/blob/main/locales/es/README.md" target="_blank">Español</a> | <a href="https://github.com/cline/cline/blob/main/locales/de/README.md" target="_blank">Deutsch</a> | <a href="https://github.com/cline/cline/blob/main/locales/ja/README.md" target="_blank">日本語</a> | <a href="https://github.com/cline/cline/blob/main/locales/zh-cn/README.md" target="_blank">简体中文</a> | <a href="https://github.com/cline/cline/blob/main/locales/zh-tw/README.md" target="_blank">繁體中文</a> | <a href="https://github.com/cline/cline/blob/main/locales/ko/README.md" target="_blank">한국어</a>
 </sub></div>
 
 # Cline – \#1 on OpenRouter
@@ -24,7 +24,7 @@ English | <a href="https://github.com/cline/cline/blob/main/locales/es/README.md
 <a href="https://github.com/cline/cline/discussions/categories/feature-requests?discussions_q=is%3Aopen+category%3A%22Feature+Requests%22+sort%3Atop" target="_blank"><strong>Feature Requests</strong></a>
 </td>
 <td align="center">
-<a href="https://cline.bot/join-us" target="_blank"><strong>We're Hiring!</strong></a>
+<a href="https://docs.cline.bot/getting-started/for-new-coders" target="_blank"><strong>Getting Started</strong></a>
 </td>
 </tbody>
 </table>
@@ -32,7 +32,7 @@ English | <a href="https://github.com/cline/cline/blob/main/locales/es/README.md
 
 Meet Cline, an AI assistant that can use your **CLI** a**N**d **E**ditor.
 
-Thanks to [Claude 3.5 Sonnet's agentic coding capabilities](https://www-cdn.anthropic.com/fed9cc193a14b84131812372d8d5857f8f304c52/Model_Card_Claude_3_Addendum.pdf), Cline can handle complex software development tasks step-by-step. With tools that let him create & edit files, explore large projects, use the browser, and execute terminal commands (after you grant permission), he can assist you in ways that go beyond code completion or tech support. Cline can even use the Model Context Protocol (MCP) to create new tools and extend his own capabilities. While autonomous AI scripts traditionally run in sandboxed environments, this extension provides a human-in-the-loop GUI to approve every file change and terminal command, providing a safe and accessible way to explore the potential of agentic AI.
+Thanks to [Claude 3.7 Sonnet's agentic coding capabilities](https://www.anthropic.com/claude/sonnet), Cline can handle complex software development tasks step-by-step. With tools that let him create & edit files, explore large projects, use the browser, and execute terminal commands (after you grant permission), he can assist you in ways that go beyond code completion or tech support. Cline can even use the Model Context Protocol (MCP) to create new tools and extend his own capabilities. While autonomous AI scripts traditionally run in sandboxed environments, this extension provides a human-in-the-loop GUI to approve every file change and terminal command, providing a safe and accessible way to explore the potential of agentic AI.
 
 1. Enter your task and add images to convert mockups into functional apps or fix bugs with screenshots.
 2. Cline starts by analyzing your file structure & source code ASTs, running regex searches, and reading relevant files to get up to speed in existing projects. By carefully managing what information is added to context, Cline can provide valuable assistance even for large, complex projects without overwhelming the context window.
@@ -187,4 +187,4 @@ To contribute to the project, start with our [Contributing Guide](CONTRIBUTING.m
 
 ## License
 
-[Apache 2.0 © 2024 Cline Bot Inc.](./LICENSE)
+[Apache 2.0 © 2025 Cline Bot Inc.](./LICENSE)

docs/PRIVACY.md

@@ -1,99 +0,0 @@
-# Cline Privacy Policy
-
-Cline Bot Inc. ("Cline," "we," "our," and/or "us") values the privacy of individuals who use our VS Code extension and related services (collectively, our "Services"). This privacy policy explains how we collect, use, and disclose information from users of our Services.
-
-## Key Points
-
--   Cline operates entirely client-side as a VS Code extension
--   No code or data is collected, stored, or transmitted to Cline's servers
--   Your data is only sent to your chosen AI provider (e.g., Anthropic, OpenAI) when you explicitly request assistance
--   All processing happens locally on your machine
--   API keys are stored securely in VS Code's built-in settings storage
-
-## Information We Process
-
-### A. Information You Provide
-
--   **API Keys**: When you choose to use certain AI model providers (OpenRouter, Anthropic, OpenAI, etc.), you provide API keys. These are stored securely and locally in your VS Code settings.
--   **Communications**: If you contact us directly (e.g., via Discord or email), we may receive information like your name, email address, and message contents.
-
-### B. Information Processing
-
-Cline functions solely as a client-side VS Code extension that facilitates communication between your editor and your chosen AI model provider:
-
-1. **File Contents**:
-
-    - Only sent to your chosen AI provider when you explicitly request assistance
-    - Never stored or transmitted to Cline's servers
-    - Only the specific files/content you select are included
-
-2. **Terminal Commands**:
-
-    - Processed entirely locally on your machine
-    - Require explicit user confirmation before execution
-    - No command history is transmitted to Cline
-
-3. **Browser Integration**:
-    - Screenshots and console logs are processed locally
-    - Temporary data is cleared after task completion
-
-## Data Security
-
-1. **Local-Only Processing**:
-
-    - All operations happen on your local machine
-    - No central servers or data collection
-    - No telemetry or usage statistics gathered
-    - No account creation required
-
-2. **API Key Security**:
-
-    - Stored using VS Code's secure settings storage system
-    - Never transmitted to Cline's servers
-    - You can remove/modify keys at any time
-
-3. **User Control**:
-    - Explicit approval required for file changes
-    - Terminal commands require confirmation
-    - Browser actions need explicit permission
-    - You control which AI provider to use
-
-## Communication with AI Providers
-
-When you request assistance:
-
-1. Selected content is sent directly to your chosen AI provider
-2. No data passes through Cline's servers
-3. Provider's own privacy policy applies to this communication:
-    - [Anthropic Privacy Policy](https://www.anthropic.com/privacy)
-    - [OpenAI Privacy Policy](https://openai.com/privacy)
-    - [OpenRouter Privacy Policy](https://openrouter.ai/privacy)
-
-## Error Handling & Debugging
-
--   Error logs are processed locally
--   No automatic error reporting to Cline
--   You control what information to include when reporting issues
-
-## Children's Privacy
-
-We do not knowingly collect, maintain, or use personal information from children under 18 years of age, and no part of our Service(s) is directed to children. If you learn that a child has provided us with personal information in violation of this Privacy Policy, then you may alert us at support@cline.bot.
-
-## Changes to Privacy Policy
-
-We will post any changes to this policy on our GitHub repository. Significant changes will be announced in our Discord community.
-
-## Security Concerns & Auditing
-
--   Cline is open source and available for security audit
--   Our client-side architecture ensures no central point of data collection
--   You can inspect exactly what data is being sent to AI providers
--   Enterprise users can implement additional access controls through VS Code
-
-## Contact Us
-
-For privacy-related questions or concerns:
-
--   Open an issue on our [GitHub repository](https://github.com/cline/cline)
--   Join our [Discord community](https://discord.gg/cline)
--   Email: support@cline.bot

docs/docs.json

@@ -0,0 +1,83 @@
+{
+	"$schema": "https://mintlify.com/docs.json",
+	"theme": "linden",
+	"name": "Cline",
+	"description": "AI-powered coding assistant for VSCode",
+	"colors": {
+		"primary": "#9D4EDD",
+		"light": "#F0E6FF",
+		"dark": "#000000"
+	},
+	"logo": {
+		"light": "/assets/robot_panel_light.png",
+		"dark": "/assets/robot_panel_dark.png"
+	},
+	"favicon": {
+		"light": "/assets/robot_panel_light.png",
+		"dark": "/assets/robot_panel_dark.png"
+	},
+	"background": {
+		"color": {
+			"light": "#F0E6FF",
+			"dark": "#000000"
+		},
+		"decoration": "gradient"
+	},
+	"styling": {
+		"eyebrows": "breadcrumbs",
+		"codeblocks": "system"
+	},
+	"appearance": {
+		"default": "system",
+		"strict": false
+	},
+	"fonts": {
+		"family": "Roboto",
+		"weight": 400
+	},
+	"navbar": {
+		"links": [
+			{
+				"label": "GitHub",
+				"href": "https://github.com/cline/cline"
+			},
+			{
+				"label": "Discord",
+				"href": "https://discord.gg/cline"
+			}
+		],
+		"primary": {
+			"type": "button",
+			"label": "Install Cline",
+			"href": "https://cline.bot/install?utm_source=website&utm_medium=header"
+		}
+	},
+	"navigation": {
+		"groups": [
+			{
+				"group": "Exploring Cline's Tools",
+				"pages": [
+					"exploring-clines-tools/cline-tools-guide",
+					"exploring-clines-tools/plan-and-act-modes-a-guide-to-effective-ai-development",
+					"exploring-clines-tools/checkpoints",
+					"exploring-clines-tools/new-task-tool",
+					"exploring-clines-tools/remote-browser-support",
+					"exploring-clines-tools/slash-commands"
+				]
+			}
+		]
+	},
+	"footer": {
+		"socials": {
+			"x": "https://x.com/cline",
+			"github": "https://github.com/cline/cline",
+			"discord": "https://discord.gg/cline"
+		}
+	},
+	"search": {
+		"prompt": "Search Cline documentation..."
+	},
+	"contextual": {
+		"options": ["copy"]
+	}
+}

docs/exploring-clines-tools/checkpoints.mdx

@@ -0,0 +1,108 @@
+---
+title: "Checkpoints and Messages"
+description: "When working with AI coding assistants, it's easy to lose control as they make rapid changes to your codebase. That's why we built Checkpoints - your safety net for experimenting confidently."
+---
+
+Checkpoints automatically save snapshots of your workspace after each step in a task. This powerful feature lets you:
+
+-   Track and review changes made during a task
+-   Roll back to any previous point if needed
+-   Experiment confidently with auto-approve mode
+-   Maintain full control over your workspace
+
+### ⚙️ How Checkpoints Work
+
+Cline creates a checkpoint after each tool use (file edits, commands, etc.). These checkpoints:
+
+-   Work alongside your Git workflow without interference
+-   Maintain context between restores
+-   Use a shadow Git repository to track changes
+
+For example, if you're working on a feature and Cline makes multiple file changes, each change creates a checkpoint. This means you can review each modification and, if needed, roll back to any point without affecting your main Git repository.
+
+#### Viewing Changes & Restoring to Checkpoint
+
+After each tool use, you can:
+
+1. Click the "Compare" button to see modified files
+2. Click the "Restore" button to open restore options
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Checkpoint comparison and restore options" />
+</Frame>
+
+#### Rolling Back
+
+To restore to a previous point:
+
+1. Click the "Restore" button next to any step
+2. Choose from three options:
+    - **Restore Task and Workspace**: Reset both codebase and task to that point
+    - **Restore Task Only**: Keep codebase changes but revert task context
+    - **Restore Workspace Only**: Reset codebase while preserving task context
+
+Example: If Cline makes changes you don't like while styling a component, you can use "Restore Workspace Only" to revert the code changes while keeping the conversation context, allowing you to try a different approach.
+
+### 💡 Use Cases
+
+Checkpoints let you be more experimental with Cline. While human coding is often methodical and iterative, AI can make substantial changes quickly. Checkpoints help you track these changes and revert if needed.
+
+#### 1. Using Auto-Approve Mode
+
+-   Provides safety net for rapid iterations
+-   Makes it easy to undo unexpected results
+
+#### 2. Testing Different Approaches
+
+-   Try multiple solutions confidently
+-   Compare different implementations
+-   Quickly revert to working states
+-   Ideal for exploring different design patterns or architectural approaches
+
+<Frame caption="In this case, I didn't like the changes Cline made to my robot dog-walking website (still working on the robots) and I wanted to revert both the codebase and the task to before any changes were made so I could start fresh.">
+	<img src="/assets/robot_panel_dark.png" alt="Checkpoint restore demo" />
+</Frame>
+
+### ✨ Best Practices
+
+1. Use checkpoints as safety nets when experimenting
+2. Leverage auto-approve mode more confidently, knowing you can always roll back
+3. Restore selectively based on needs:
+
+    - Use "Restore Task and Workspace" for a fresh start, reversing changes to files and the task conversation.
+    - Use "Restore Task Only" to try different prompts, but leave all files as they exist
+    - Use "Restore Workspace Only" to attempt different implementations, or prune context from the task
+
+🛟 Checkpoints are your safety net when working with Cline, enabling you to experiment freely while maintaining full control over your codebase. Whether you're refactoring a complex component, trying different implementation approaches, or using auto-approve mode for rapid development, checkpoints ensure you can always review changes and roll back if needed.
+
+#### 🗑️ Deleting Checkpoints
+
+You can delete all checkpoints by using the **"Delete All History"** button in the task history menu. Note that this will also delete all tasks. Checkpoints are stored in VS Code's globalStorage.
+
+---
+
+## Editing Messages
+
+Cline allows you to edit chat messages in a task after they've been submitted (with the exception of the message that started the task).
+
+Perhaps you didn't get the results you wanted, thought of a better way to phrase your request, or need to add more information. Editing your message allows you to re-submit a request without starting over or restoring your files or workspace with checkpoints. There are two Restore options:
+
+-   **"Restore Chat"** restores just the task state and re-submits an API request to your provider with your edited message.
+
+-   **"Restore All"** restores both the task state and workspace state before re-submitting an API request. "Workspace state" refers to the condition of your workspace (files, content, etc.) at different points in the conversation.
+
+**Interactive Editing:**
+
+-   Messages can be clicked to enter edit mode
+-   Cline automatically selects all text when entering edit mode
+
+**Keyboard Shortcuts:**
+
+-   Escape: Exit edit mode
+-   Enter: Restore just the task
+-   Cmd/Ctrl + Enter: Restore the task and workspace
+-   Shift + Enter: Insert new line / line break
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Message editing interface" />
+</Frame>

docs/exploring-clines-tools/cline-tools-guide.mdx

@@ -0,0 +1,139 @@
+---
+title: "Cline Tools Reference Guide"
+---
+
+## What Can Cline Do?
+
+Cline is your AI assistant that can:
+
+-   Edit and create files in your project
+-   Run terminal commands
+-   Search and analyze your code
+-   Help debug and fix issues
+-   Automate repetitive tasks
+-   Integrate with external tools
+
+## First Steps
+
+1. **Start a Task**
+
+    - Type your request in the chat
+    - Example: "Create a new React component called Header"
+
+2. **Provide Context**
+
+    - Use @ mentions to add files, folders, or URLs
+    - Example: "@file:src/components/App.tsx"
+
+3. **Review Changes**
+    - Cline will show diffs before making changes
+    - You can edit or reject changes
+
+## Key Features
+
+1. **File Editing**
+
+    - Create new files
+    - Modify existing code
+    - Search and replace across files
+
+2. **Terminal Commands**
+
+    - Run npm commands
+    - Start development servers
+    - Install dependencies
+
+3. **Code Analysis**
+
+    - Find and fix errors
+    - Refactor code
+    - Add documentation
+
+4. **Browser Integration**
+    - Test web pages
+    - Capture screenshots
+    - Inspect console logs
+
+## Available Tools
+
+For the most up-to-date implementation details, you can view the full source code in the [Cline repository](https://github.com/cline/cline/blob/main/src/core/Cline.ts).
+
+Cline has access to the following tools for various tasks:
+
+1. **File Operations**
+
+    - `write_to_file`: Create or overwrite files
+    - `read_file`: Read file contents
+    - `replace_in_file`: Make targeted edits to files
+    - `search_files`: Search files using regex
+    - `list_files`: List directory contents
+
+2. **Terminal Operations**
+
+    - `execute_command`: Run CLI commands
+    - `list_code_definition_names`: List code definitions
+
+3. **MCP Tools**
+
+    - `use_mcp_tool`: Use tools from MCP servers
+    - `access_mcp_resource`: Access MCP server resources
+    - Users can create custom MCP tools that Cline can then access
+    - Example: Create a weather API tool that Cline can use to fetch forecasts
+
+4. **Interaction Tools**
+    - `ask_followup_question`: Ask user for clarification
+    - `attempt_completion`: Present final results
+
+Each tool has specific parameters and usage patterns. Here are some examples:
+
+-   Create a new file (write_to_file):
+
+    ```xml
+    <write_to_file>
+    <path>src/components/Header.tsx</path>
+    <content>
+    // Header component code
+    </content>
+    </write_to_file>
+    ```
+
+-   Search for a pattern (search_files):
+
+    ```xml
+    <search_files>
+    <path>src</path>
+    <regex>function\s+\w+\(</regex>
+    <file_pattern>*.ts</file_pattern>
+    </search_files>
+    ```
+
+-   Run a command (execute_command):
+    ```xml
+    <execute_command>
+    <command>npm install axios</command>
+    <requires_approval>false</requires_approval>
+    </execute_command>
+    ```
+
+## Common Tasks
+
+1. **Create a New Component**
+
+    - "Create a new React component called Footer"
+
+2. **Fix a Bug**
+
+    - "Fix the error in src/utils/format.ts"
+
+3. **Refactor Code**
+
+    - "Refactor the Button component to use TypeScript"
+
+4. **Run Commands**
+    - "Run npm install to add axios"
+
+## Getting Help
+
+-   [Join the Discord community](https://discord.gg/cline)
+-   Check the documentation
+-   Provide feedback to improve Cline

docs/exploring-clines-tools/new-task-tool.mdx

@@ -0,0 +1,381 @@
+---
+title: "New Task Tool"
+---
+
+### The `new_task` Tool & Context Management Strategies
+
+#### Overview
+
+Cline includes a powerful internal tool, `new_task`, designed to help manage workflow continuity and context preservation, especially during complex or long-running tasks. This tool, combined with Cline's awareness of its own context window usage and the flexibility of `.clinerules`, enables sophisticated strategies for breaking down work and ensuring seamless transitions between task sessions.
+
+Understanding the core capabilities and how they interact with custom rules is key to leveraging this feature effectively.
+
+#### Core Capabilities
+
+Two fundamental capabilities enable advanced context management:
+
+1. **The `new_task` Tool:**
+    - **Function:** Allows Cline, upon user approval, to end the current task session and immediately start a new one.
+    - **Context Preloading:** Crucially, Cline can **preload** this new task session with specific context provided within the tool's `<context>` block. This context can be anything Cline or a `.clinerules` file defines – summaries, code snippets, next steps, project state, etc.
+2. **Context Window Awareness:**
+    - **Tracking:** Cline internally tracks the percentage of its available context window currently being used during a task.
+    - **Visibility:** This information is visible in the `environment_details` provided to Cline in its prompt.
+
+#### Using the `/newtask` Slash Command
+
+As a quick alternative to Cline suggesting the `newtask` tool or defining complex rules, you can directly initiate the process using a Slash Command.
+
+-   **How:** Simply type `/newtask` in the chat input field.
+-   **Action:** Cline will propose creating a new task, typically suggesting context based on the current session (similar to its default behavior when using the tool). You will still get the `ask_followup_question` prompt to confirm and potentially modify the context before the new task is created.
+-   **Benefit:** Provides a fast, user-initiated way to leverage the `new_task` functionality for branching explorations or managing long sessions without waiting for Cline to suggest it.
+
+#### Default Behavior (Without `.clinerules`)
+
+By default, without specific `.clinerules` dictating its behavior:
+
+-   **Tool Availability:** The `new_task` tool exists, and Cline _can_ choose to use it.
+-   **Context Awareness:** Cline _is_ aware of its context usage percentage.
+-   **No Automatic Trigger:** Cline **will not** automatically initiate a task handoff _solely_ based on context usage reaching a specific percentage (like 50%). The decision to suggest using `new_task` comes from the AI model's reasoning based on the overall task progress and prompt instructions.
+-   **Basic Context Preloading:** If `new_task` is used without specific rules defining the `<context>` block structure, Cline will attempt to preload relevant information based on its current understanding (e.g., a basic summary of progress and next steps), but this may be less comprehensive than a rule-driven approach.
+
+#### The Power of `.clinerules`: Enabling Custom Workflows
+
+While the core capabilities exist by default, the true power, automation, and customization emerge when you combine `new_task` and context awareness with custom workflows defined in `.clinerules`. This allows you to precisely control _when_ and _how_ Cline manages context and task continuity.
+
+Key benefits of using `.clinerules` with `new_task`:
+
+-   **Automated Context Management:** Define rules to automatically trigger handoffs at specific context percentages (e.g., >50%, >70%) or token counts, ensuring optimal performance and preventing context loss.
+-   **Model-Specific Optimization:** Tailor handoff triggers based on known thresholds for different LLMs (e.g., trigger earlier for models known to degrade past a certain token count).
+-   **Intelligent Breakpoints:** Instruct Cline via rules to find logical stopping points (e.g., after completing a function or test) _after_ a context threshold is passed, ensuring cleaner handoffs.
+-   **Structured Task Decomposition:** Use Plan Mode to define subtasks, then use `.clinerules` to have Cline automatically create a new task via `new_task` upon completing each subtask, preloading the context for the _next_ subtask.
+-   **Custom Context Packaging:** Mandate the exact structure and content of the `<context>` block in `.clinerules` for highly detailed and consistent handoffs (see example below).
+-   **Improved Memory Persistence:** Use `new_task` context blocks as a primary, integrated way to persist information across sessions, potentially replacing or supplementing file-based memory systems.
+-   **Workflow Automation:** Define rules for specific scenarios, like always preloading certain setup instructions or project boilerplate when starting tasks of a particular type.
+
+#### Example Rule-Driven Workflow: Task Handoff Process
+
+A common workflow, **driven by specific `.clinerules` like the example below**, involves these steps:
+
+1. **Trigger Identification (Rule-Based):** Cline monitors for handoff points defined in the rules (e.g., context usage > 50%, task completion).
+2. **User Confirmation:** Cline uses `ask_followup_question` to propose creating a new task, often showing the intended context defined by the rules.
+
+    ```xml
+    <ask_followup_question>
+    <question>I've completed [specific accomplishment] and context usage is high (XX%). Would you like me to create a new task to continue with [remaining work], preloading the following context?</question>
+    <options>["Yes, create new task", "Modify context first", "No, continue this session"]</options>
+    </ask_followup_question>
+    ```
+
+3. **User Control:** You can approve, deny, or ask Cline to modify the context before the new task is created.
+4. **Context Packaging (`new_task` Tool):** If approved, Cline uses `new_task`, packaging the context according to the structure mandated by the `.clinerules`.
+5. **New Task Creation:** The current task ends, and a new session begins immediately, preloaded with the specified context.
+
+#### The Handoff Context Block (Rule-Defined Structure)
+
+The effectiveness of rule-driven handoffs depends heavily on how `.clinerules` define the `<context>` block. A comprehensive structure often includes:
+
+-   **`## Completed Work`**: Detailed list of accomplishments, files modified/created, key decisions.
+-   **`## Current State`**: Project status, running processes, key file states.
+-   **`## Next Steps`**: Clear, prioritized list of remaining tasks, implementation details, known challenges.
+-   **`## Reference Information`**: Links, code snippets, patterns, user preferences.
+-   **Actionable Start:** A clear instruction for the immediate next action.
+
+#### Potential Use Cases & Workflows
+
+The flexibility of `new_task` combined with `.clinerules` opens up many possibilities:
+
+-   **Proactive Context Window Management:** Automatically trigger handoffs at specific percentages (e.g., 50%, 70%) or token counts to maintain optimal performance.
+-   **Intelligent Breakpoints:** Instruct Cline to find logical stopping points (e.g., after completing a function or test) _after_ a context threshold is passed, ensuring cleaner handoffs.
+-   **Structured Task Decomposition:** Use Plan Mode to define subtasks, then use `.clinerules` to have Cline automatically create a new task via `new_task` upon completing each subtask.
+-   **Automated Session Summaries:** Configure the `<context>` block to always include a summary of the previous session's key discussion points.
+-   **Preloading Boilerplate/Setup:** Start new tasks related to specific projects preloaded with standard setup instructions or file templates.
+-   **"Memory Bank" Alternative:** Use `new_task` context blocks as the primary way to persist information across sessions, potentially replacing file-based memory systems.
+
+Experimenting with `.clinerules` is encouraged to discover workflows that best suit your needs!
+
+#### Example `.clinerules`: Task Handoff Strategy Guide
+
+Below is an example `.clinerules` file focused specifically on using `new_task` for context window management. **Remember, this is just one specific strategy; the core `new_task` tool can be used differently with other custom rules.**
+
+````markdown
+# You MUST use the `new_task` tool: Task Handoff Strategy Guide
+
+**⚠️ CRITICAL INSTRUCTIONS - YOU MUST FOLLOW THESE GUIDELINES ⚠️**
+
+This guide provides **MANDATORY** instructions for effectively breaking down complex tasks and implementing a smooth handoff process between tasks. You **MUST** follow these guidelines to ensure continuity, context preservation, and efficient task completion.
+
+## ⚠️ CONTEXT WINDOW MONITORING - MANDATORY ACTION REQUIRED ⚠️
+
+You **MUST** monitor the context window usage displayed in the environment details. When usage exceeds 50% of the available context window, you **MUST** initiate a task handoff using the `new_task` tool.
+
+Example of context window usage over 50% with a 200K context window:
+
+\`\`\`text
+
+# Context Window Usage
+
+105,000 / 200,000 tokens (53%)
+Model: anthropic/claude-3.7-sonnet (200K context window)
+\`\`\`
+
+**IMPORTANT**: When you see context window usage at or above 50%, you MUST:
+
+1. Complete your current logical step
+2. Use the `ask_followup_question` tool to offer creating a new task
+3. If approved, use the `new_task` tool with comprehensive handoff instructions
+
+## Task Breakdown in Plan Mode - REQUIRED PROCESS
+
+Plan Mode is specifically designed for analyzing complex tasks and breaking them into manageable subtasks. When in Plan Mode, you **MUST**:
+
+### 1. Initial Task Analysis - REQUIRED
+
+-   **MUST** begin by thoroughly understanding the full scope of the user's request
+-   **MUST** identify all major components and dependencies of the task
+-   **MUST** consider potential challenges, edge cases, and prerequisites
+
+### 2. Strategic Task Decomposition - REQUIRED
+
+-   **MUST** break the overall task into logical, discrete subtasks
+-   **MUST** prioritize subtasks based on dependencies (what must be completed first)
+-   **MUST** aim for subtasks that can be completed within a single session (15-30 minutes of work)
+-   **MUST** consider natural breaking points where context switching makes sense
+
+### 3. Creating a Task Roadmap - REQUIRED
+
+-   **MUST** present a clear, numbered list of subtasks to the user
+-   **MUST** explain dependencies between subtasks
+-   **MUST** provide time estimates for each subtask when possible
+-   **MUST** use Mermaid diagrams to visualize task flow and dependencies when helpful
+
+\`\`\`mermaid
+graph TD
+A[Main Task] --> B[Subtask 1: Setup]
+A --> C[Subtask 2: Core Implementation]
+A --> D[Subtask 3: Testing]
+A --> E[Subtask 4: Documentation]
+B --> C
+C --> D
+\`\`\`
+
+### 4. Getting User Approval - REQUIRED
+
+-   **MUST** ask for user feedback on the proposed task breakdown
+-   **MUST** adjust the plan based on user priorities or additional requirements
+-   **MUST** confirm which subtask to begin with
+-   **MUST** request the user to toggle to Act Mode when ready to implement
+
+## Task Implementation and Handoff Process - MANDATORY PROCEDURES
+
+When implementing tasks in Act Mode, you **MUST** follow these guidelines for effective task handoff:
+
+### 1. Focused Implementation - REQUIRED
+
+-   **MUST** focus on completing the current subtask fully
+-   **MUST** document progress clearly through comments and commit messages
+-   **MUST** create checkpoints at logical completion points
+
+### 2. Recognizing Completion Points - CRITICAL
+
+You **MUST** identify natural handoff points when:
+
+-   The current subtask is fully completed
+-   You've reached a logical stopping point in a larger subtask
+-   The implementation is taking longer than expected and can be continued later
+-   The task scope has expanded beyond the original plan
+-   **CRITICAL**: The context window usage exceeds 50% (e.g., 100,000+ tokens for a 200K context window)
+
+### 3. Initiating the Handoff Process - MANDATORY ACTION
+
+When you've reached a completion point, you **MUST**:
+
+1. Summarize what has been accomplished so far
+2. Clearly state what remains to be done
+3. **MANDATORY**: Use the `ask_followup_question` tool to offer creating a new task:
+
+\`\`\`xml
+<ask_followup_question>
+<question>I've completed [specific accomplishment]. Would you like me to create a new task to continue with [remaining work]?</question>
+<options>["Yes, create a new task", "No, continue in this session", "Let me think about it"]</options>
+</ask_followup_question>
+\`\`\`
+
+### 4. Creating a New Task with Context - REQUIRED ACTION
+
+If the user agrees to create a new task, you **MUST** use the `new_task` tool with comprehensive handoff instructions:
+
+\`\`\`xml
+<new_task>
+<context>
+
+# Task Continuation: [Brief Task Title]
+
+## Completed Work
+
+-   [Detailed list of completed items]
+-   [Include specific files modified/created]
+-   [Note any important decisions made]
+
+## Current State
+
+-   [Description of the current state of the project]
+-   [Any running processes or environment setup]
+-   [Key files and their current state]
+
+## Next Steps
+
+-   [Detailed list of remaining tasks]
+-   [Specific implementation details to address]
+-   [Any known challenges to be aware of]
+
+## Reference Information
+
+-   [Links to relevant documentation]
+-   [Important code snippets or patterns to follow]
+-   [Any user preferences noted during the current session]
+
+Please continue the implementation by [specific next action].
+</context>
+</new_task>
+\`\`\`
+
+### 5. Detailed Context Transfer - MANDATORY COMPONENTS
+
+When creating a new task, you **MUST** always include:
+
+#### Project Context - REQUIRED
+
+-   **MUST** include the overall goal and purpose of the project
+-   **MUST** include key architectural decisions and patterns
+-   **MUST** include technology stack and dependencies
+
+#### Implementation Details - REQUIRED
+
+-   **MUST** list files created or modified in the current session
+-   **MUST** describe specific functions, classes, or components implemented
+-   **MUST** explain design patterns being followed
+-   **MUST** outline testing approach
+
+#### Progress Tracking - REQUIRED
+
+-   **MUST** provide checklist of completed items
+-   **MUST** provide checklist of remaining items
+-   **MUST** note any blockers or challenges encountered
+
+#### User Preferences - REQUIRED
+
+-   **MUST** note coding style preferences mentioned by the user
+-   **MUST** document specific approaches requested by the user
+-   **MUST** highlight priority areas identified by the user
+
+## Best Practices for Effective Handoffs - MANDATORY GUIDELINES
+
+### 1. Maintain Continuity - REQUIRED
+
+-   **MUST** use consistent terminology between tasks
+-   **MUST** reference previous decisions and their rationale
+-   **MUST** maintain the same architectural approach unless explicitly changing direction
+
+### 2. Preserve Context - REQUIRED
+
+-   **MUST** include relevant code snippets in the handoff
+-   **MUST** summarize key discussions from the previous session
+-   **MUST** reference specific files and line numbers when applicable
+
+### 3. Set Clear Next Actions - REQUIRED
+
+-   **MUST** begin the handoff with a clear, actionable next step
+-   **MUST** prioritize remaining tasks
+-   **MUST** highlight any decisions that need to be made
+
+### 4. Document Assumptions - REQUIRED
+
+-   **MUST** clearly state any assumptions made during implementation
+-   **MUST** note areas where user input might be needed
+-   **MUST** identify potential alternative approaches
+
+### 5. Optimize for Resumability - REQUIRED
+
+-   **MUST** structure the handoff so the next session can begin working immediately
+-   **MUST** include setup instructions if environment configuration is needed
+-   **MUST** provide a quick summary at the top for rapid context restoration
+
+## Example Task Handoff
+
+### Example #1 of an effective task handoff:
+
+\`\`\`xml
+<new_task>
+<context>
+
+# Task Continuation: Implement User Authentication System
+
+## Completed Work
+
+-   Created basic Express.js server structure
+-   Implemented MongoDB connection and user schema
+-   Completed user registration endpoint with password hashing
+-   Added input validation using Joi
+-   Created initial test suite for registration endpoint
+
+## Current State
+
+-   Server runs successfully on port 3000
+-   MongoDB connection is established
+-   Registration endpoint (/api/users/register) is fully functional
+-   Test suite passes for all registration scenarios
+
+## Next Steps
+
+1. Implement login endpoint (/api/users/login)
+    - Use bcrypt to compare passwords
+    - Generate JWT token upon successful login
+    - Add proper error handling for invalid credentials
+2. Create authentication middleware
+    - Verify JWT tokens
+    - Extract user information
+    - Handle expired tokens
+3. Add protected routes that require authentication
+4. Implement password reset functionality
+
+## Reference Information
+
+-   JWT secret should be stored in .env file
+-   Follow the existing error handling pattern in routes/users.js
+-   User schema is defined in models/User.js
+-   Test patterns are established in tests/auth.test.js
+
+Please continue by implementing the login endpoint following the same patterns established in the registration endpoint.
+</context>
+</new_task>
+\`\`\`
+
+### Example #2 of an ineffective task handoff:
+
+_(Note: The example provided in the original rules showing "YOLO MODE Implementation" seems less like a direct handoff context block and more like a general status update with future considerations. A true ineffective handoff might lack detail in 'Current State' or 'Next Steps')._
+
+## When to Use Task Handoffs - MANDATORY TRIGGERS
+
+You **MUST** initiate task handoffs in these scenarios:
+
+1. **CRITICAL**: When context window usage exceeds 50% (e.g., 100,000+ tokens for a 200K context window)
+2. **Long-running projects** that exceed a single session
+3. **Complex implementations** with multiple distinct phases
+4. **When context window limitations** are approaching
+5. **When switching focus areas** within a larger project
+6. **When different expertise** might be beneficial for different parts of the task
+
+**⚠️ FINAL REMINDER - CRITICAL INSTRUCTION ⚠️**
+
+You **MUST** monitor the context window usage in the environment details section. When it exceeds 50% (e.g., "105,000 / 200,000 tokens (53%)"), you **MUST** proactively initiate the task handoff process using the `ask_followup_question` tool followed by the `new_task` tool. You MUST use the `new_task` tool.
+
+By strictly following these guidelines, you'll ensure smooth transitions between tasks, maintain project momentum, and provide the best possible experience for users working on complex, multi-session projects.
+
+```markdown
+## User Interaction & Workflow Considerations
+
+-   **Linear Flow:** Currently, using `new_task` creates a linear sequence. The old task ends, and the new one begins. The old task history remains accessible for backtracking.
+-   **User Approval:** You always have control, approving the handoff and having the chance to modify the context Cline proposes to carry forward.
+-   **Flexibility:** The core `new_task` tool is a flexible building block. Experiment with `.clinerules` to create workflows that best suit your needs, whether for strict context management, task decomposition, or other creative uses.
+```
+````

docs/exploring-clines-tools/plan-and-act-modes-a-guide-to-effective-ai-development.mdx

@@ -0,0 +1,136 @@
+---
+title: "Plan & Act Modes: A Guide to Effective AI Development"
+---
+
+## Overview
+
+Plan & Act modes represent Cline's approach to structured AI development, emphasizing thoughtful planning before implementation. This dual-mode system helps developers create more maintainable, accurate code while reducing iteration time.
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Use Plan to gather context before using Act to implement the plan" />
+</Frame>
+
+### Understanding the Modes
+
+#### Plan Mode
+
+-   Optimized for context gathering and strategy
+-   Cannot make changes to your codebase
+-   Focused on understanding requirements and creating implementation plans
+-   Enables full file reading for comprehensive project understanding
+
+#### Act Mode
+
+-   Streamlined for implementation based on established plans
+-   Has access to all of Cline's building capabilities
+-   Maintains context from the planning phase
+-   Can execute changes to your codebase
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Act mode capabilities" />
+</Frame>
+
+### Workflow Guide
+
+#### 1. Start with Plan Mode
+
+Begin every significant development task in Plan mode:
+
+In this mode:
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Plan mode workflow" />
+</Frame>
+
+-   Share your requirements
+-   Let Cline analyze relevant files
+-   Engage in dialogue to clarify objectives
+-   Develop implementation strategy
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Planning phase" />
+</Frame>
+
+#### 2. Switch to Act Mode
+
+Once you have a clear plan, switch to Act mode:
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Switching to Act mode" />
+</Frame>
+
+Act mode allows Cline to:
+
+-   Execute against the agreed plan
+-   Make changes to your codebase
+-   Maintain context from planning phase
+
+#### 3. Iterate as Needed
+
+Complex projects often require multiple plan-act cycles:
+
+-   Return to Plan mode when encountering unexpected complexity
+-   Use Act mode for implementing solutions
+-   Maintain development momentum while ensuring quality
+
+### Best Practices
+
+#### Planning Phase
+
+1. Be comprehensive with requirements
+2. Share relevant context upfront
+3. Point Cline to relevant files if he hasn't read them
+4. Validate approach before implementation
+
+#### Implementation Phase
+
+1. Follow the established plan
+2. Monitor progress against objectives
+3. Track changes and their impact
+4. Document significant decisions
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Implementation best practices" />
+</Frame>
+
+### Power User Tips
+
+#### Enhancing Planning
+
+-   Use Plan mode to explore edge cases before implementation
+-   Switch back to Plan when encountering unexpected complexity
+-   Leverage file reading to validate assumptions early
+-   Have Cline write markdown files of the plan for future reference
+
+### Common Patterns
+
+#### When to Use Plan Mode
+
+-   Starting new features
+-   Debugging complex issues
+-   Architectural decisions
+-   Requirements analysis
+
+#### When to Use Act Mode
+
+-   Implementing agreed solutions
+-   Making routine changes
+-   Following established patterns
+-   Executing test cases
+
+<Frame>
+	<img src="/assets/robot_panel_dark.png" alt="Mode usage patterns" />
+</Frame>
+
+### Contributing
+
+Share your experiences and improvements:
+
+-   Join our [Discord community](https://discord.gg/cline)
+-   Participate in discussions
+-   Submit feature requests
+-   Report issues
+
+---
+
+Remember: The time invested in planning pays dividends in implementation quality and maintenance efficiency

docs/exploring-clines-tools/remote-browser-support.mdx

@@ -0,0 +1,114 @@
+---
+title: "Remote Browser Support"
+description: "Remote browser support allows Cline to utilize a remote Chrome instance, leveraging authentication tokens and session cookies relevant to certain web development test cases."
+icon: globe-pointer
+---
+
+The Remote Browser feature in Cline allows the AI assistant to interact with web content directly through a controlled browser instance. This enables several powerful capabilities:
+
+-   Viewing and interacting with websites
+-   Testing locally running web applications
+-   Monitoring console logs and errors
+-   Performing browser actions like clicking, typing, and scrolling
+
+## Remote Browser in Cline
+
+### What is Remote Browser?
+
+Remote Browser allows Cline to view and interact with websites directly. This feature enables Cline to:
+
+-   Visit websites and view their content
+-   Test your locally running web applications
+-   Fill out forms and click on elements
+-   Capture screenshots of what it sees
+-   Scroll through pages to see more content
+
+### How to Use Remote Browser
+
+#### Basic Commands
+
+You can ask Cline to use the browser with simple instructions:
+
+-   **Open a website**: "Use the browser to check the website at [https://example.com](https://example.com/)"
+-   **Click on elements**: "Click the login button"
+-   **Type text**: "Type 'Hello world' in the search box"
+-   **Scroll the page**: "Scroll down to see more content"
+-   **Close the browser**: "Close the browser now"
+
+#### Example Workflows
+
+**Testing a Web Application:**
+
+```javascript
+Can you start my React app with "npm start" and then check if it's working properly at http://localhost:3000?
+```
+
+**Analyzing a Website:**
+
+```javascript
+Can you visit https://example.com and tell me what you think about its design and layout?
+```
+
+**Filling Out a Form:**
+
+```javascript
+Please go to https://example.com/contact, fill out the contact form with some test data, and submit it.
+```
+
+### Important Things to Know
+
+#### One Browser at a Time
+
+Cline can only use one browser at a time. If you want to visit a different website, you can either:
+
+-   Ask Cline to navigate to a new URL within the same browser session
+-   Ask Cline to close the current browser and open a new one
+
+#### Browser Must Be Closed Before Using Other Tools
+
+If you want Cline to edit files or run commands after using the browser, you must first ask it to close the browser:
+
+```javascript
+Close the browser and then update the CSS file to fix the alignment issue we saw.
+```
+
+#### What Cline Sees
+
+The browser has a fixed viewport size (900x600 pixels by default), similar to a small laptop screen. Cline will share screenshots after each action so you can see exactly what it sees.
+
+#### Console Logs
+
+Cline captures browser console logs, which can be helpful for debugging web applications. These logs are included with each screenshot.
+
+### Common Use Cases
+
+-   **Web Development**: Test your websites and web applications
+-   **UI/UX Review**: Get feedback on website design and usability
+-   **Content Research**: Have Cline browse websites to gather information
+-   **Form Testing**: Verify that forms work correctly
+-   **Responsive Design Testing**: Check how websites look at different screen sizes
+
+### Troubleshooting
+
+-   **If a website doesn't load**: Try providing a direct URL with the http:// or https:// prefix
+-   **If clicking doesn't work**: Try describing the location of the element more precisely
+-   **If the browser seems stuck**: Ask Cline to close the browser and try again
+
+### Using Remote Browser with VS Code in WSL
+
+When running VS Code in WSL, you'll need to configure Windows to allow WSL to connect to Chrome. Follow these steps:
+
+#### Open PowerShell as Administrator and Run:
+
+```powershell
+# Allow WSL to connect to Chrome's debugging port
+New-NetFirewallRule -DisplayName "WSL Chrome Debug" -Direction Inbound -LocalPort 9222 -Protocol TCP -Action Allow
+```
+
+#### Configure Cline in VS Code:
+
+1. Open VS Code settings
+2. Search for "Cline: Chrome Executable Path"
+3. Set the value to the path of your Chrome executable (e.g., `C:\Program Files\Google\Chrome\Application\chrome.exe`)
+
+Cline should now be able to use the Remote Browser feature from within WSL.

docs/exploring-clines-tools/slash-commands.mdx

@@ -0,0 +1,57 @@
+---
+title: "Slash Commands"
+---
+
+#### Overview
+
+Cline provides slash commands as a quick way to invoke specific tools or actions directly from the chat input, offering shortcuts for common operations. This page details the available slash commands and their usage.
+
+#### /newtask
+
+The `/newtask` slash command provides a fast, user-initiated way to leverage the `new_task` tool's functionality for branching explorations or managing long sessions without waiting for Cline to suggest it.
+
+**Functionality:**
+
+1. **Initiation:** Typing `/newtask` in the chat input signals Cline to prepare for starting a new task session.
+2. **Context Proposal:** Cline proposes creating a new task and typically suggests context to preload based on the current session (summarizing key aspects like current work, technical concepts, relevant files, problems solved, and next steps).
+3. **User Confirmation:** You will receive a confirmation prompt (via the `ask_followup_question` tool) displaying the proposed context. You can approve it directly or modify the context before the new task begins.
+4. **New Session:** Upon confirmation, Cline ends the current task session and immediately starts a new one, preloaded with the approved context.
+
+**Benefit:** Allows you to cleanly branch your work or start a new phase while carrying over essential background information ("knowledge transfer") without manual copying or losing the thread of the previous session.
+
+#### /smol (alias /compact)
+
+The `/smol` slash command (with `/compact` as an alias) allows you to condense the chat history **within your current task**. This is useful when a conversation becomes very long, potentially impacting performance or making it harder for the model to maintain focus.
+
+**Functionality:**
+
+1. **Initiation:** Typing `/smol` or `/compact` tells Cline you want to condense the current chat history. You can optionally add instructions after the command to guide the summarization process (e.g., `/smol focus only on the database changes` or `/smol be concise, use bullet points`).
+2. **Summarization:** Cline analyzes the conversation history, considering any additional instructions provided, and generates a summary focusing on key elements: recent discussion points, important decisions, technical concepts, relevant files, problems solved, and planned next steps. Cline determines the appropriate length and detail for the summary. It retains the beginning and very recent parts of the chat while summarizing the middle sections.
+3. **User Confirmation:** Cline presents this generated summary to you via a confirmation prompt and asks if it accurately reflects the essential context.
+4. **Condensing:** If you approve the summary, Cline replaces the summarized middle portion of the chat history in its active context with the generated summary. This reduces the overall token count for subsequent interactions within the _same task_.
+5. **Feedback:** If you reject the summary or provide feedback, Cline will retain the original history and incorporate your feedback for future actions.
+
+**Benefit:** Helps maintain focus and manage token usage during very long, continuous tasks (like deep debugging or extended feature development) without needing to start an entirely new task session. Allows user guidance on the summarization focus.
+
+#### When to Use Which?
+
+Choosing between `/newtask` and `/smol` depends on your goal:
+
+-   Use `/smol` (or `/compact`) when:
+    -   You want to continue the **same task**, but the chat history has become very long or costly.
+    -   You need to reduce token usage for upcoming interactions within the current workflow.
+    -   Example: Deep debugging session where you want to summarize previous steps before continuing.
+-   Use `/newtask` when:
+    -   You have finished one phase of work and want to start a **fresh, related task**.
+    -   You want to branch your exploration while preserving key context from the previous session.
+    -   Example: Moving from developing Feature A to starting work on Feature B, carrying over relevant architectural decisions.
+
+#### Why Manage Context?
+
+While Cline supports large context windows, actively managing context using tools and commands like `/newtask` and `/smol` is often beneficial:
+
+-   **Performance:** Large language models can sometimes experience performance degradation or lose focus when context windows become extremely full (e.g., over 50-75% capacity, depending on the model). Condensing or resetting context can help maintain optimal performance.
+-   **Relevance:** Summarizing or starting fresh ensures the most relevant information is prioritized in the context window.
+-   **Cost:** Reducing the number of tokens sent to the model in each turn can help manage costs, especially with more expensive models.
+
+Using `/newtask` and `/smol` provides you with direct control over the conversation context, allowing for more efficient and effective interaction with Cline.

esbuild.js

@@ -8,6 +8,60 @@ const watch = process.argv.includes("--watch")
 /**
  * @type {import('esbuild').Plugin}
  */
+const aliasResolverPlugin = {
+	name: "alias-resolver",
+	setup(build) {
+		const aliases = {
+			"@": path.resolve(__dirname, "src"),
+			"@api": path.resolve(__dirname, "src/api"),
+			"@core": path.resolve(__dirname, "src/core"),
+			"@integrations": path.resolve(__dirname, "src/integrations"),
+			"@services": path.resolve(__dirname, "src/services"),
+			"@shared": path.resolve(__dirname, "src/shared"),
+			"@utils": path.resolve(__dirname, "src/utils"),
+			"@packages": path.resolve(__dirname, "src/packages"),
+		}
+
+		// For each alias entry, create a resolver
+		Object.entries(aliases).forEach(([alias, aliasPath]) => {
+			const aliasRegex = new RegExp(`^${alias}($|/.*)`)
+			build.onResolve({ filter: aliasRegex }, (args) => {
+				const importPath = args.path.replace(alias, aliasPath)
+
+				// First, check if the path exists as is
+				if (fs.existsSync(importPath)) {
+					const stats = fs.statSync(importPath)
+					if (stats.isDirectory()) {
+						// If it's a directory, try to find index files
+						const extensions = [".ts", ".tsx", ".js", ".jsx"]
+						for (const ext of extensions) {
+							const indexFile = path.join(importPath, `index${ext}`)
+							if (fs.existsSync(indexFile)) {
+								return { path: indexFile }
+							}
+						}
+					} else {
+						// It's a file that exists, so return it
+						return { path: importPath }
+					}
+				}
+
+				// If the path doesn't exist, try appending extensions
+				const extensions = [".ts", ".tsx", ".js", ".jsx"]
+				for (const ext of extensions) {
+					const pathWithExtension = `${importPath}${ext}`
+					if (fs.existsSync(pathWithExtension)) {
+						return { path: pathWithExtension }
+					}
+				}
+
+				// If nothing worked, return the original path and let esbuild handle the error
+				return { path: importPath }
+			})
+		})
+	},
+}
+
 const esbuildProblemMatcherPlugin = {
 	name: "esbuild-problem-matcher",
 
@@ -52,6 +106,7 @@ const copyWasmFiles = {
 				"java",
 				"php",
 				"swift",
+				"kotlin",
 			]
 
 			languages.forEach((lang) => {
@@ -67,10 +122,23 @@ const extensionConfig = {
 	minify: production,
 	sourcemap: !production,
 	logLevel: "silent",
+	define: {
+		"process.env.IS_DEV": JSON.stringify(!production),
+	},
+	tsconfig: path.resolve(__dirname, "tsconfig.json"),
 	plugins: [
 		copyWasmFiles,
+		aliasResolverPlugin,
 		/* add to the end of plugins array */
 		esbuildProblemMatcherPlugin,
+		{
+			name: "alias-plugin",
+			setup(build) {
+				build.onResolve({ filter: /^pkce-challenge$/ }, (args) => {
+					return { path: require.resolve("pkce-challenge/dist/index.browser.js") }
+				})
+			},
+		},
 	],
 	entryPoints: ["src/extension.ts"],
 	format: "cjs",

evals/.gitignore

@@ -0,0 +1,3 @@
+repositories
+
+results/evals.db

evals/README.md

@@ -0,0 +1,186 @@
+# Cline Evaluation System
+
+This directory contains the evaluation system for benchmarking Cline against various coding evaluation frameworks.
+
+## Overview
+
+The Cline Evaluation System allows you to:
+
+1. Run Cline against standardized coding benchmarks
+2. Collect comprehensive metrics on performance
+3. Generate detailed reports on evaluation results
+4. Compare performance across different models and benchmarks
+
+## Architecture
+
+The evaluation system consists of two main components:
+
+1. **Test Server**: Enhanced HTTP server in `src/services/test/TestServer.ts` that provides detailed task results
+2. **CLI Tool**: Command-line interface in `evals/cli/` for orchestrating evaluations
+
+## Directory Structure
+
+```
+cline-repo/
+├── src/
+│   ├── services/
+│   │   ├── test/
+│   │   │   ├── TestServer.ts         # Enhanced HTTP server for task execution
+│   │   │   ├── GitHelper.ts          # Git utilities for file tracking
+│   │   │   └── ...
+│   │   └── ...
+│   └── ...
+├── evals/                            # Main directory for evaluation system
+│   ├── cli/                          # CLI tool for orchestrating evaluations
+│   │   ├── src/
+│   │   │   ├── index.ts              # CLI entry point
+│   │   │   ├── commands/             # CLI commands (setup, run, report)
+│   │   │   ├── adapters/             # Benchmark adapters
+│   │   │   ├── db/                   # Database management
+│   │   │   └── utils/                # Utility functions
+│   │   ├── package.json
+│   │   └── tsconfig.json
+│   ├── repositories/                 # Cloned benchmark repositories
+│   │   ├── exercism/                 # Modified Exercism (from pashpashpash/evals)
+│   │   ├── swe-bench/                # SWE-Bench repository
+│   │   ├── swelancer/                # SWELancer repository
+│   │   └── multi-swe/                # Multi-SWE-Bench repository
+│   ├── results/                      # Evaluation results storage
+│   │   ├── runs/                     # Individual run results
+│   │   └── reports/                  # Generated reports
+│   └── README.md                     # This file
+└── ...
+```
+
+## Getting Started
+
+### Prerequisites
+
+- Node.js 16+
+- VSCode with Cline extension installed
+- Git
+
+### Activation Mechanism
+
+The evaluation system uses an `evals.env` file approach to activate test mode in the Cline extension. When an evaluation is run:
+
+1. The CLI creates an `evals.env` file in the workspace directory
+2. The Cline extension activates due to the `workspaceContains:evals.env` activation event
+3. The extension detects this file and automatically enters test mode
+4. After evaluation completes, the file is automatically removed
+
+This approach eliminates the need for environment variables during the build process and allows for targeted activation only when needed for evaluations. The extension remains dormant during normal use, only activating when an evals.env file is present. For more details, see [Evals Env Activation](./docs/evals-env-activation.md).
+
+### Installation
+
+1. Build the CLI tool:
+
+```bash
+cd evals/cli
+npm install
+npm run build
+```
+
+### Usage
+
+#### Setting Up Benchmarks
+
+```bash
+cd evals/cli
+node dist/index.js setup
+```
+
+This will clone and set up all benchmark repositories. You can specify specific benchmarks:
+
+```bash
+node dist/index.js setup --benchmarks exercism
+```
+
+#### Running Evaluations
+
+```bash
+node dist/index.js run --model claude-3-opus-20240229 --benchmark exercism
+```
+
+Options:
+- `--model`: The model to evaluate (default: claude-3-opus-20240229)
+- `--benchmark`: Specific benchmark to run (default: all)
+- `--count`: Number of tasks to run (default: all)
+
+#### Generating Reports
+
+```bash
+node dist/index.js report
+```
+
+Options:
+- `--format`: Report format (json, markdown) (default: markdown)
+- `--output`: Output path for the report
+
+#### Managing Test Mode Activation
+
+The CLI provides a command to manually manage the evals.env file for test mode activation:
+
+```bash
+node dist/index.js evals-env create  # Create evals.env file in current directory
+node dist/index.js evals-env remove  # Remove evals.env file from current directory
+node dist/index.js evals-env check   # Check if evals.env file exists in current directory
+```
+
+Options:
+- `--directory`: Specify a directory other than the current one
+
+## Benchmarks
+
+### Exercism
+
+Modified Exercism exercises from the [pashpashpash/evals](https://github.com/pashpashpash/evals) repository. These are small, focused programming exercises in various languages.
+
+### SWE-Bench (Coming Soon)
+
+Real-world software engineering tasks from the [SWE-bench](https://github.com/SWE-bench/SWE-bench) repository.
+
+### SWELancer (Coming Soon)
+
+Freelance-style programming tasks from the SWELancer benchmark.
+
+### Multi-SWE-Bench (Coming Soon)
+
+Multi-file software engineering tasks from the Multi-SWE-Bench repository.
+
+## Metrics
+
+The evaluation system collects the following metrics:
+
+- **Token Usage**: Input and output tokens
+- **Cost**: Estimated cost of API calls
+- **Duration**: Time taken to complete tasks
+- **Tool Usage**: Number of tool calls and failures
+- **Success Rate**: Percentage of tasks completed successfully
+- **Functional Correctness**: Percentage of tests passed
+
+## Reports
+
+Reports are generated in Markdown or JSON format and include:
+
+- Overall summary
+- Benchmark-specific results
+- Model-specific results
+- Tool usage statistics
+- Charts and visualizations
+
+## Development
+
+### Adding a New Benchmark
+
+1. Create a new adapter in `evals/cli/src/adapters/`
+2. Implement the `BenchmarkAdapter` interface
+3. Register the adapter in `evals/cli/src/adapters/index.ts`
+
+### Extending Metrics
+
+To add new metrics:
+
+1. Update the database schema in `evals/cli/src/db/schema.ts`
+2. Add collection logic in `evals/cli/src/utils/results.ts`
+3. Update report generation in `evals/cli/src/commands/report.ts`

evals/cli/package.json

@@ -0,0 +1,39 @@
+{
+	"name": "cline-evaluation-cli",
+	"version": "0.1.0",
+	"description": "CLI tool for orchestrating Cline evaluations across multiple benchmarks",
+	"main": "dist/index.js",
+	"scripts": {
+		"build": "tsc",
+		"start": "node dist/index.js",
+		"dev": "ts-node src/index.ts",
+		"test": "echo \"Error: no test specified\" && exit 1"
+	},
+	"keywords": [
+		"cline",
+		"evaluation",
+		"benchmark"
+	],
+	"author": "",
+	"license": "MIT",
+	"dependencies": {
+		"better-sqlite3": "^8.0.0",
+		"chalk": "^4.1.2",
+		"commander": "^9.4.1",
+		"execa": "^5.1.1",
+		"node-fetch": "^2.7.0",
+		"ora": "^5.4.1",
+		"sqlite": "^4.1.2",
+		"uuid": "^9.0.0",
+		"yargs": "^17.6.2"
+	},
+	"devDependencies": {
+		"@types/better-sqlite3": "^7.6.3",
+		"@types/node": "^18.11.18",
+		"@types/node-fetch": "^2.6.12",
+		"@types/uuid": "^9.0.0",
+		"@types/yargs": "^17.0.19",
+		"ts-node": "^10.9.1",
+		"typescript": "^4.9.4"
+	}
+}

evals/cli/src/adapters/exercism.ts

@@ -0,0 +1,194 @@
+import * as path from "path"
+import * as fs from "fs"
+import execa from "execa"
+import { BenchmarkAdapter, Task, VerificationResult } from "./types"
+
+const EVALS_DIR = path.resolve(__dirname, "../../../")
+
+/**
+ * Adapter for the modified Exercism benchmark
+ */
+export class ExercismAdapter implements BenchmarkAdapter {
+	name = "exercism"
+
+	/**
+	 * Set up the Exercism benchmark repository
+	 */
+	async setup(): Promise<void> {
+		// Clone repository if needed
+		const exercismDir = path.join(EVALS_DIR, "repositories", "exercism")
+
+		if (!fs.existsSync(exercismDir)) {
+			console.log(`Cloning Exercism repository to ${exercismDir}...`)
+			await execa("git", ["clone", "https://github.com/pashpashpash/evals.git", exercismDir])
+			console.log("Exercism repository cloned successfully")
+		} else {
+			console.log(`Exercism repository already exists at ${exercismDir}`)
+
+			// Pull latest changes
+			console.log("Pulling latest changes...")
+			await execa("git", ["pull"], { cwd: exercismDir })
+			console.log("Repository updated successfully")
+		}
+	}
+
+	/**
+	 * List all available tasks in the Exercism benchmark
+	 */
+	async listTasks(): Promise<Task[]> {
+		const tasks: Task[] = []
+		const exercisesDir = path.join(EVALS_DIR, "repositories", "exercism")
+
+		// Ensure the repository exists
+		if (!fs.existsSync(exercisesDir)) {
+			throw new Error(`Exercism repository not found at ${exercisesDir}. Run setup first.`)
+		}
+
+		// Read language directories
+		const languages = fs
+			.readdirSync(exercisesDir)
+			.filter((dir) => fs.statSync(path.join(exercisesDir, dir)).isDirectory())
+			.filter((dir) => !dir.startsWith(".") && !["node_modules", ".git"].includes(dir))
+
+		for (const language of languages) {
+			const languageDir = path.join(exercisesDir, language)
+
+			// Read exercise directories
+			const exercises = fs.readdirSync(languageDir).filter((dir) => fs.statSync(path.join(languageDir, dir)).isDirectory())
+
+			for (const exercise of exercises) {
+				const exerciseDir = path.join(languageDir, exercise)
+
+				// Read instructions
+				let description = ""
+				const instructionsPath = path.join(exerciseDir, "docs", "instructions.md")
+				if (fs.existsSync(instructionsPath)) {
+					description = fs.readFileSync(instructionsPath, "utf-8")
+				}
+
+				// Determine test commands based on language
+				let testCommands: string[] = []
+				switch (language) {
+					case "javascript":
+						testCommands = ["npm install", "npm test"]
+						break
+					case "python":
+						testCommands = ["python -m pytest -o markers=task *_test.py"]
+						break
+					case "go":
+						testCommands = ["go test"]
+						break
+					case "java":
+						testCommands = ["./gradlew test"]
+						break
+					case "rust":
+						testCommands = ["cargo test"]
+						break
+					default:
+						testCommands = []
+				}
+
+				tasks.push({
+					id: `exercism-${language}-${exercise}`,
+					name: exercise,
+					description,
+					workspacePath: exerciseDir,
+					setupCommands: [],
+					verificationCommands: testCommands,
+					metadata: {
+						language,
+						type: "exercism",
+					},
+				})
+			}
+		}
+
+		return tasks
+	}
+
+	/**
+	 * Prepare a specific task for execution
+	 * @param taskId The ID of the task to prepare
+	 */
+	async prepareTask(taskId: string): Promise<Task> {
+		const tasks = await this.listTasks()
+		const task = tasks.find((t) => t.id === taskId)
+
+		if (!task) {
+			throw new Error(`Task ${taskId} not found`)
+		}
+
+		// Check if Git repository is already initialized
+		const gitDirExists = fs.existsSync(path.join(task.workspacePath, ".git"))
+
+		try {
+			// Initialize Git repository if needed
+			if (!gitDirExists) {
+				await execa("git", ["init"], { cwd: task.workspacePath })
+			}
+
+			// Create a dummy file to ensure there's something to commit
+			const dummyFilePath = path.join(task.workspacePath, ".eval-timestamp")
+			fs.writeFileSync(dummyFilePath, new Date().toISOString())
+
+			// Add all files and commit
+			await execa("git", ["add", "."], { cwd: task.workspacePath })
+
+			try {
+				await execa("git", ["commit", "-m", "Initial commit"], { cwd: task.workspacePath })
+			} catch (error: any) {
+				// If commit fails because there are no changes, that's okay
+				if (!error.stderr?.includes("nothing to commit")) {
+					throw error
+				}
+			}
+		} catch (error: any) {
+			console.warn(`Warning: Git operations failed: ${error.message}`)
+			console.warn("Continuing without Git initialization")
+		}
+
+		return task
+	}
+
+	/**
+	 * Verify the result of a task execution
+	 * @param task The task that was executed
+	 * @param result The result of the task execution
+	 */
+	async verifyResult(task: Task, result: any): Promise<VerificationResult> {
+		// Run verification commands
+		let success = true
+		let output = ""
+
+		for (const command of task.verificationCommands) {
+			try {
+				const [cmd, ...args] = command.split(" ")
+				const { stdout } = await execa(cmd, args, { cwd: task.workspacePath })
+				output += stdout + "\n"
+			} catch (error: any) {
+				success = false
+				if (error.stdout) {
+					output += error.stdout + "\n"
+				}
+				if (error.stderr) {
+					output += error.stderr + "\n"
+				}
+			}
+		}
+
+		// Parse test results
+		const testsPassed = (output.match(/PASS/g) || []).length
+		const testsFailed = (output.match(/FAIL/g) || []).length
+		const testsTotal = testsPassed + testsFailed
+
+		return {
+			success,
+			metrics: {
+				testsPassed,
+				testsFailed,
+				testsTotal,
+				functionalCorrectness: testsTotal > 0 ? testsPassed / testsTotal : 0,
+			},
+		}
+	}
+}

evals/cli/src/adapters/index.ts

@@ -0,0 +1,47 @@
+import { BenchmarkAdapter } from "./types"
+import { ExercismAdapter } from "./exercism"
+import { SWEBenchAdapter } from "./swe-bench"
+import { SWELancerAdapter } from "./swelancer"
+import { MultiSWEAdapter } from "./multi-swe"
+
+// Registry of all available adapters
+const adapters: Record<string, BenchmarkAdapter> = {
+	// Exercism is the primary adapter with real implementation
+	exercism: new ExercismAdapter(),
+
+	// Dummy adapters for testing
+	"swe-bench": new SWEBenchAdapter(),
+	swelancer: new SWELancerAdapter(),
+	"multi-swe": new MultiSWEAdapter(),
+}
+
+/**
+ * Get a specific adapter by name
+ * @param name The name of the adapter to get
+ * @returns The requested adapter
+ * @throws Error if the adapter is not found
+ */
+export function getAdapter(name: string): BenchmarkAdapter {
+	const adapter = adapters[name]
+	if (!adapter) {
+		throw new Error(`Adapter for benchmark '${name}' not found`)
+	}
+	return adapter
+}
+
+/**
+ * Get all available adapters
+ * @returns Array of all registered adapters
+ */
+export function getAllAdapters(): BenchmarkAdapter[] {
+	return Object.values(adapters)
+}
+
+/**
+ * Register a new adapter
+ * @param name The name to register the adapter under
+ * @param adapter The adapter to register
+ */
+export function registerAdapter(name: string, adapter: BenchmarkAdapter): void {
+	adapters[name] = adapter
+}

evals/cli/src/adapters/multi-swe.ts

@@ -0,0 +1,192 @@
+import * as path from "path"
+import * as fs from "fs"
+import execa from "execa"
+import { BenchmarkAdapter, Task, VerificationResult } from "./types"
+
+const EVALS_DIR = path.resolve(__dirname, "../../../")
+
+/**
+ * Dummy adapter for the Multi-SWE-Bench benchmark
+ */
+export class MultiSWEAdapter implements BenchmarkAdapter {
+	name = "multi-swe"
+
+	/**
+	 * Set up the Multi-SWE-Bench benchmark repository (dummy implementation)
+	 */
+	async setup(): Promise<void> {
+		console.log("Multi-SWE-Bench dummy setup completed")
+
+		// Create repositories directory if it doesn't exist
+		const repoDir = path.join(EVALS_DIR, "repositories", "multi-swe")
+		if (!fs.existsSync(repoDir)) {
+			fs.mkdirSync(repoDir, { recursive: true })
+			console.log(`Created dummy Multi-SWE-Bench directory at ${repoDir}`)
+		}
+	}
+
+	/**
+	 * List all available tasks in the Multi-SWE-Bench benchmark (dummy implementation)
+	 */
+	async listTasks(): Promise<Task[]> {
+		return [
+			{
+				id: "multi-swe-task-1",
+				name: "Multi-Language API Integration",
+				description:
+					"Implement a system that integrates a Python backend with a TypeScript frontend and a Rust processing service.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "multi-swe"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					languages: ["python", "typescript", "rust"],
+					complexity: "high",
+					type: "multi-swe",
+				},
+			},
+			{
+				id: "multi-swe-task-2",
+				name: "Cross-Platform Mobile App",
+				description: "Create a cross-platform mobile app using React Native with native modules in Swift and Kotlin.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "multi-swe"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					languages: ["javascript", "swift", "kotlin"],
+					complexity: "medium",
+					type: "multi-swe",
+				},
+			},
+			{
+				id: "multi-swe-task-3",
+				name: "Microservice Architecture",
+				description: "Design and implement a microservice architecture with services written in Go, Node.js, and Java.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "multi-swe"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					languages: ["go", "javascript", "java"],
+					complexity: "high",
+					type: "multi-swe",
+				},
+			},
+		]
+	}
+
+	/**
+	 * Prepare a specific task for execution (dummy implementation)
+	 * @param taskId The ID of the task to prepare
+	 */
+	async prepareTask(taskId: string): Promise<Task> {
+		const tasks = await this.listTasks()
+		const task = tasks.find((t) => t.id === taskId)
+
+		if (!task) {
+			throw new Error(`Task ${taskId} not found`)
+		}
+
+		// Create a dummy workspace for the task
+		const taskDir = path.join(task.workspacePath, taskId)
+		if (!fs.existsSync(taskDir)) {
+			fs.mkdirSync(taskDir, { recursive: true })
+
+			// Create a dummy file for the task
+			fs.writeFileSync(
+				path.join(taskDir, "README.md"),
+				`# ${task.name}\n\n${task.description}\n\nThis is a dummy task for testing purposes.`,
+			)
+
+			// Create additional dummy files based on task type
+			if (task.id === "multi-swe-task-1") {
+				// Python backend
+				fs.mkdirSync(path.join(taskDir, "backend"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "backend", "app.py"),
+					`# TODO: Implement Python backend\nfrom flask import Flask\n\napp = Flask(__name__)\n\n@app.route('/')\ndef hello():\n    return "Hello, World!"\n`,
+				)
+
+				// TypeScript frontend
+				fs.mkdirSync(path.join(taskDir, "frontend"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "frontend", "app.ts"),
+					`// TODO: Implement TypeScript frontend\nconsole.log('Frontend starting...');\n`,
+				)
+
+				// Rust processing service
+				fs.mkdirSync(path.join(taskDir, "processor"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "processor", "main.rs"),
+					`// TODO: Implement Rust processing service\nfn main() {\n    println!("Processor starting...");\n}\n`,
+				)
+			} else if (task.id === "multi-swe-task-2") {
+				// React Native app
+				fs.mkdirSync(path.join(taskDir, "app"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "app", "App.js"),
+					`// TODO: Implement React Native app\nimport React from 'react';\nimport { View, Text } from 'react-native';\n\nexport default function App() {\n  return (\n    <View>\n      <Text>Hello, World!</Text>\n    </View>\n  );\n}\n`,
+				)
+
+				// Swift native module
+				fs.mkdirSync(path.join(taskDir, "ios"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "ios", "NativeModule.swift"),
+					`// TODO: Implement Swift native module\nimport Foundation\n\n@objc(NativeModule)\nclass NativeModule: NSObject {\n  @objc\n  func hello() -> String {\n    return "Hello from Swift"\n  }\n}\n`,
+				)
+
+				// Kotlin native module
+				fs.mkdirSync(path.join(taskDir, "android"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "android", "NativeModule.kt"),
+					`// TODO: Implement Kotlin native module\npackage com.example.app\n\nclass NativeModule {\n  fun hello(): String {\n    return "Hello from Kotlin"\n  }\n}\n`,
+				)
+			} else if (task.id === "multi-swe-task-3") {
+				// Go service
+				fs.mkdirSync(path.join(taskDir, "service-go"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "service-go", "main.go"),
+					`// TODO: Implement Go service\npackage main\n\nimport "fmt"\n\nfunc main() {\n\tfmt.Println("Go service starting...")\n}\n`,
+				)
+
+				// Node.js service
+				fs.mkdirSync(path.join(taskDir, "service-node"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "service-node", "server.js"),
+					`// TODO: Implement Node.js service\nconsole.log('Node.js service starting...');\n`,
+				)
+
+				// Java service
+				fs.mkdirSync(path.join(taskDir, "service-java"), { recursive: true })
+				fs.writeFileSync(
+					path.join(taskDir, "service-java", "Main.java"),
+					`// TODO: Implement Java service\npublic class Main {\n    public static void main(String[] args) {\n        System.out.println("Java service starting...");\n    }\n}\n`,
+				)
+			}
+		}
+
+		// Update the task's workspace path to the task-specific directory
+		return {
+			...task,
+			workspacePath: taskDir,
+		}
+	}
+
+	/**
+	 * Verify the result of a task execution (dummy implementation)
+	 * @param task The task that was executed
+	 * @param result The result of the task execution
+	 */
+	async verifyResult(task: Task, result: any): Promise<VerificationResult> {
+		// Always return success for dummy implementation
+		return {
+			success: true,
+			metrics: {
+				testsPassed: 1,
+				testsFailed: 0,
+				testsTotal: 1,
+				functionalCorrectness: 1.0,
+				crossLanguageIntegration: 0.9, // Dummy metric specific to Multi-SWE
+				architectureQuality: 0.85, // Dummy metric specific to Multi-SWE
+			},
+		}
+	}
+}

evals/cli/src/adapters/swe-bench.ts

@@ -0,0 +1,125 @@
+import * as path from "path"
+import * as fs from "fs"
+import execa from "execa"
+import { BenchmarkAdapter, Task, VerificationResult } from "./types"
+
+const EVALS_DIR = path.resolve(__dirname, "../../../")
+
+/**
+ * Dummy adapter for the SWE-Bench benchmark
+ */
+export class SWEBenchAdapter implements BenchmarkAdapter {
+	name = "swe-bench"
+
+	/**
+	 * Set up the SWE-Bench benchmark repository (dummy implementation)
+	 */
+	async setup(): Promise<void> {
+		console.log("SWE-Bench dummy setup completed")
+
+		// Create repositories directory if it doesn't exist
+		const repoDir = path.join(EVALS_DIR, "repositories", "swe-bench")
+		if (!fs.existsSync(repoDir)) {
+			fs.mkdirSync(repoDir, { recursive: true })
+			console.log(`Created dummy SWE-Bench directory at ${repoDir}`)
+		}
+	}
+
+	/**
+	 * List all available tasks in the SWE-Bench benchmark (dummy implementation)
+	 */
+	async listTasks(): Promise<Task[]> {
+		return [
+			{
+				id: "swe-bench-task-1",
+				name: "Fix React Component Bug",
+				description: "Fix a bug in a React component where the state is not properly updated.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "swe-bench"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					repository: "facebook/react",
+					issue: "#12345",
+					type: "swe-bench",
+				},
+			},
+			{
+				id: "swe-bench-task-2",
+				name: "Optimize Database Query",
+				description: "Optimize a slow database query in a Django application.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "swe-bench"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					repository: "django/django",
+					issue: "#6789",
+					type: "swe-bench",
+				},
+			},
+			{
+				id: "swe-bench-task-3",
+				name: "Fix Memory Leak",
+				description: "Fix a memory leak in a Node.js application.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "swe-bench"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					repository: "nodejs/node",
+					issue: "#9876",
+					type: "swe-bench",
+				},
+			},
+		]
+	}
+
+	/**
+	 * Prepare a specific task for execution (dummy implementation)
+	 * @param taskId The ID of the task to prepare
+	 */
+	async prepareTask(taskId: string): Promise<Task> {
+		const tasks = await this.listTasks()
+		const task = tasks.find((t) => t.id === taskId)
+
+		if (!task) {
+			throw new Error(`Task ${taskId} not found`)
+		}
+
+		// Create a dummy workspace for the task
+		const taskDir = path.join(task.workspacePath, taskId)
+		if (!fs.existsSync(taskDir)) {
+			fs.mkdirSync(taskDir, { recursive: true })
+
+			// Create a dummy file for the task
+			fs.writeFileSync(
+				path.join(taskDir, "README.md"),
+				`# ${task.name}\n\n${task.description}\n\nThis is a dummy task for testing purposes.`,
+			)
+		}
+
+		// Update the task's workspace path to the task-specific directory
+		return {
+			...task,
+			workspacePath: taskDir,
+		}
+	}
+
+	/**
+	 * Verify the result of a task execution (dummy implementation)
+	 * @param task The task that was executed
+	 * @param result The result of the task execution
+	 */
+	async verifyResult(task: Task, result: any): Promise<VerificationResult> {
+		// Always return success for dummy implementation
+		return {
+			success: true,
+			metrics: {
+				testsPassed: 1,
+				testsFailed: 0,
+				testsTotal: 1,
+				functionalCorrectness: 1.0,
+				performanceImprovement: 0.25, // Dummy metric specific to SWE-Bench
+				codeQuality: 0.9, // Dummy metric specific to SWE-Bench
+			},
+		}
+	}
+}

evals/cli/src/adapters/swelancer.ts

@@ -0,0 +1,143 @@
+import * as path from "path"
+import * as fs from "fs"
+import execa from "execa"
+import { BenchmarkAdapter, Task, VerificationResult } from "./types"
+
+const EVALS_DIR = path.resolve(__dirname, "../../../")
+
+/**
+ * Dummy adapter for the SWELancer benchmark
+ */
+export class SWELancerAdapter implements BenchmarkAdapter {
+	name = "swelancer"
+
+	/**
+	 * Set up the SWELancer benchmark repository (dummy implementation)
+	 */
+	async setup(): Promise<void> {
+		console.log("SWELancer dummy setup completed")
+
+		// Create repositories directory if it doesn't exist
+		const repoDir = path.join(EVALS_DIR, "repositories", "swelancer")
+		if (!fs.existsSync(repoDir)) {
+			fs.mkdirSync(repoDir, { recursive: true })
+			console.log(`Created dummy SWELancer directory at ${repoDir}`)
+		}
+	}
+
+	/**
+	 * List all available tasks in the SWELancer benchmark (dummy implementation)
+	 */
+	async listTasks(): Promise<Task[]> {
+		return [
+			{
+				id: "swelancer-task-1",
+				name: "Create Landing Page",
+				description: "Create a responsive landing page for a new product using HTML, CSS, and JavaScript.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "swelancer"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					client: "TechStartup Inc.",
+					difficulty: "medium",
+					type: "swelancer",
+				},
+			},
+			{
+				id: "swelancer-task-2",
+				name: "Build REST API",
+				description: "Create a RESTful API for a blog application using Node.js and Express.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "swelancer"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					client: "BlogCo",
+					difficulty: "hard",
+					type: "swelancer",
+				},
+			},
+			{
+				id: "swelancer-task-3",
+				name: "Fix CSS Layout Issues",
+				description: "Fix layout issues in a responsive website across different screen sizes.",
+				workspacePath: path.join(EVALS_DIR, "repositories", "swelancer"),
+				setupCommands: [],
+				verificationCommands: [],
+				metadata: {
+					client: "DesignAgency",
+					difficulty: "easy",
+					type: "swelancer",
+				},
+			},
+		]
+	}
+
+	/**
+	 * Prepare a specific task for execution (dummy implementation)
+	 * @param taskId The ID of the task to prepare
+	 */
+	async prepareTask(taskId: string): Promise<Task> {
+		const tasks = await this.listTasks()
+		const task = tasks.find((t) => t.id === taskId)
+
+		if (!task) {
+			throw new Error(`Task ${taskId} not found`)
+		}
+
+		// Create a dummy workspace for the task
+		const taskDir = path.join(task.workspacePath, taskId)
+		if (!fs.existsSync(taskDir)) {
+			fs.mkdirSync(taskDir, { recursive: true })
+
+			// Create a dummy file for the task
+			fs.writeFileSync(
+				path.join(taskDir, "README.md"),
+				`# ${task.name}\n\n${task.description}\n\nThis is a dummy task for testing purposes.`,
+			)
+
+			// Create additional dummy files based on task type
+			if (task.id === "swelancer-task-1") {
+				fs.writeFileSync(
+					path.join(taskDir, "index.html"),
+					`<!DOCTYPE html>\n<html>\n<head>\n  <title>Landing Page</title>\n</head>\n<body>\n  <!-- TODO: Implement landing page -->\n</body>\n</html>`,
+				)
+			} else if (task.id === "swelancer-task-2") {
+				fs.writeFileSync(
+					path.join(taskDir, "server.js"),
+					`// TODO: Implement REST API\nconsole.log('Server starting...');`,
+				)
+			} else if (task.id === "swelancer-task-3") {
+				fs.writeFileSync(
+					path.join(taskDir, "styles.css"),
+					`/* TODO: Fix layout issues */\nbody {\n  margin: 0;\n  padding: 0;\n}`,
+				)
+			}
+		}
+
+		// Update the task's workspace path to the task-specific directory
+		return {
+			...task,
+			workspacePath: taskDir,
+		}
+	}
+
+	/**
+	 * Verify the result of a task execution (dummy implementation)
+	 * @param task The task that was executed
+	 * @param result The result of the task execution
+	 */
+	async verifyResult(task: Task, result: any): Promise<VerificationResult> {
+		// Always return success for dummy implementation
+		return {
+			success: true,
+			metrics: {
+				testsPassed: 1,
+				testsFailed: 0,
+				testsTotal: 1,
+				functionalCorrectness: 1.0,
+				clientSatisfaction: 0.95, // Dummy metric specific to SWELancer
+				timeEfficiency: 0.85, // Dummy metric specific to SWELancer
+			},
+		}
+	}
+}

evals/cli/src/adapters/types.ts

@@ -0,0 +1,31 @@
+/**
+ * Represents a task to be executed
+ */
+export interface Task {
+	id: string
+	name: string
+	description: string
+	workspacePath: string
+	setupCommands: string[]
+	verificationCommands: string[]
+	metadata: Record<string, any>
+}
+
+/**
+ * Result of verifying a task execution
+ */
+export interface VerificationResult {
+	success: boolean
+	metrics: Record<string, any>
+}
+
+/**
+ * Interface for benchmark adapters
+ */
+export interface BenchmarkAdapter {
+	name: string
+	setup(): Promise<void>
+	listTasks(): Promise<Task[]>
+	prepareTask(taskId: string): Promise<Task>
+	verifyResult(task: Task, result: any): Promise<VerificationResult>
+}

evals/cli/src/commands/evals-env.ts

@@ -0,0 +1,53 @@
+import * as path from "path"
+import chalk from "chalk"
+import { createEvalsEnvFile, removeEvalsEnvFile, checkEvalsEnvFile } from "../utils/evals-env"
+
+interface EvalsEnvOptions {
+	action: "create" | "remove" | "check"
+	directory?: string
+}
+
+/**
+ * Handler for the evals-env command
+ * @param options Command options
+ */
+export async function evalsEnvHandler(options: EvalsEnvOptions): Promise<void> {
+	// Determine the directory to use - default to repository root instead of current directory
+	const currentDir = process.cwd()
+	const repoRoot = path.resolve(currentDir, "..", "..") // Navigate up from evals/cli to root
+	const directory = options.directory || repoRoot
+
+	console.log(chalk.blue(`Working with directory: ${directory}`))
+
+	// Perform the requested action
+	switch (options.action) {
+		case "create":
+			console.log(chalk.blue("Creating evals.env file..."))
+			createEvalsEnvFile(directory)
+			console.log(chalk.green("The Cline extension should now detect this file and enter test mode."))
+			console.log(chalk.yellow("Note: You may need to reload VSCode for the changes to take effect."))
+			break
+
+		case "remove":
+			console.log(chalk.blue("Removing evals.env file..."))
+			removeEvalsEnvFile(directory)
+			console.log(chalk.green("The Cline extension should now exit test mode."))
+			console.log(chalk.yellow("Note: You may need to reload VSCode for the changes to take effect."))
+			break
+
+		case "check":
+			console.log(chalk.blue("Checking for evals.env file..."))
+			const exists = checkEvalsEnvFile(directory)
+			if (exists) {
+				console.log(chalk.green("The Cline extension should be in test mode."))
+			} else {
+				console.log(chalk.yellow("The Cline extension should not be in test mode."))
+			}
+			break
+
+		default:
+			console.error(chalk.red(`Unknown action: ${options.action}`))
+			console.log(chalk.yellow("Valid actions are: create, remove, check"))
+			break
+	}
+}

evals/cli/src/commands/report.ts

@@ -0,0 +1,237 @@
+import * as fs from "fs"
+import * as path from "path"
+import chalk from "chalk"
+import ora from "ora"
+import { ResultsDatabase } from "../db"
+import { generateMarkdownReport } from "../utils/markdown"
+
+interface ReportOptions {
+	format?: "json" | "markdown"
+	output?: string
+}
+
+/**
+ * Handler for the report command
+ * @param options Command options
+ */
+export async function reportHandler(options: ReportOptions): Promise<void> {
+	const format = options.format || "markdown"
+	const db = new ResultsDatabase()
+
+	try {
+		const spinner = ora("Generating report...").start()
+
+		// Get all runs
+		const runs = db.getRuns()
+
+		console.log(chalk.blue(`Found ${runs.length} evaluation runs`))
+
+		if (runs.length === 0) {
+			spinner.fail("No evaluation runs found")
+			return
+		}
+
+		// Generate summary report
+		const summary = {
+			runs: runs.length,
+			models: [...new Set(runs.map((run) => run.model))],
+			benchmarks: [...new Set(runs.map((run) => run.benchmark))],
+			tasks: 0,
+			successRate: 0,
+			averageTokens: 0,
+			averageCost: 0,
+			averageDuration: 0,
+			totalToolCalls: 0,
+			totalToolFailures: 0,
+			toolSuccessRate: 0,
+			toolUsage: {} as Record<string, { calls: number; failures: number }>,
+		}
+
+		let totalTasks = 0
+		let successfulTasks = 0
+		let totalTokens = 0
+		let totalCost = 0
+		let totalDuration = 0
+		let totalToolCalls = 0
+		let totalToolFailures = 0
+
+		for (const run of runs) {
+			const tasks = db.getRunTasks(run.id)
+			totalTasks += tasks.length
+
+			for (const task of tasks) {
+				if (task.success) {
+					successfulTasks++
+				}
+
+				const metrics = db.getTaskMetrics(task.id)
+
+				const tokensIn = metrics.find((m) => m.name === "tokensIn")?.value || 0
+				const tokensOut = metrics.find((m) => m.name === "tokensOut")?.value || 0
+				totalTokens += tokensIn + tokensOut
+
+				totalCost += metrics.find((m) => m.name === "cost")?.value || 0
+				totalDuration += metrics.find((m) => m.name === "duration")?.value || 0
+
+				// Collect tool call metrics
+				totalToolCalls += task.total_tool_calls || 0
+				totalToolFailures += task.total_tool_failures || 0
+
+				// Get detailed tool usage
+				const toolCalls = db.getTaskToolCalls(task.id)
+
+				for (const toolCall of toolCalls) {
+					if (!summary.toolUsage[toolCall.tool_name]) {
+						summary.toolUsage[toolCall.tool_name] = {
+							calls: 0,
+							failures: 0,
+						}
+					}
+
+					summary.toolUsage[toolCall.tool_name].calls += toolCall.call_count
+					summary.toolUsage[toolCall.tool_name].failures += toolCall.failure_count
+				}
+			}
+		}
+
+		// Calculate tool success rate
+		summary.totalToolCalls = totalToolCalls
+		summary.totalToolFailures = totalToolFailures
+		summary.toolSuccessRate = totalToolCalls > 0 ? 1 - totalToolFailures / totalToolCalls : 1.0
+
+		summary.tasks = totalTasks
+		summary.successRate = totalTasks > 0 ? successfulTasks / totalTasks : 0
+		summary.averageTokens = totalTasks > 0 ? totalTokens / totalTasks : 0
+		summary.averageCost = totalTasks > 0 ? totalCost / totalTasks : 0
+		summary.averageDuration = totalTasks > 0 ? totalDuration / totalTasks : 0
+
+		// Generate benchmark-specific reports
+		const benchmarkReports: Record<string, any> = {}
+
+		for (const benchmark of summary.benchmarks) {
+			const benchmarkRuns = runs.filter((run) => run.benchmark === benchmark)
+			const benchmarkSummary = {
+				runs: benchmarkRuns.length,
+				models: [...new Set(benchmarkRuns.map((run) => run.model))],
+				tasks: 0,
+				successRate: 0,
+				averageTokens: 0,
+				averageCost: 0,
+				averageDuration: 0,
+			}
+
+			let benchmarkTasks = 0
+			let benchmarkSuccessfulTasks = 0
+			let benchmarkTotalTokens = 0
+			let benchmarkTotalCost = 0
+			let benchmarkTotalDuration = 0
+
+			for (const run of benchmarkRuns) {
+				const tasks = db.getRunTasks(run.id)
+				benchmarkTasks += tasks.length
+
+				for (const task of tasks) {
+					if (task.success) {
+						benchmarkSuccessfulTasks++
+					}
+
+					const metrics = db.getTaskMetrics(task.id)
+
+					const tokensIn = metrics.find((m) => m.name === "tokensIn")?.value || 0
+					const tokensOut = metrics.find((m) => m.name === "tokensOut")?.value || 0
+					benchmarkTotalTokens += tokensIn + tokensOut
+
+					benchmarkTotalCost += metrics.find((m) => m.name === "cost")?.value || 0
+					benchmarkTotalDuration += metrics.find((m) => m.name === "duration")?.value || 0
+				}
+			}
+
+			benchmarkSummary.tasks = benchmarkTasks
+			benchmarkSummary.successRate = benchmarkTasks > 0 ? benchmarkSuccessfulTasks / benchmarkTasks : 0
+			benchmarkSummary.averageTokens = benchmarkTasks > 0 ? benchmarkTotalTokens / benchmarkTasks : 0
+			benchmarkSummary.averageCost = benchmarkTasks > 0 ? benchmarkTotalCost / benchmarkTasks : 0
+			benchmarkSummary.averageDuration = benchmarkTasks > 0 ? benchmarkTotalDuration / benchmarkTasks : 0
+
+			benchmarkReports[benchmark] = benchmarkSummary
+		}
+
+		// Generate model-specific reports
+		const modelReports: Record<string, any> = {}
+
+		for (const model of summary.models) {
+			const modelRuns = runs.filter((run) => run.model === model)
+			const modelSummary = {
+				runs: modelRuns.length,
+				benchmarks: [...new Set(modelRuns.map((run) => run.benchmark))],
+				tasks: 0,
+				successRate: 0,
+				averageTokens: 0,
+				averageCost: 0,
+				averageDuration: 0,
+			}
+
+			let modelTasks = 0
+			let modelSuccessfulTasks = 0
+			let modelTotalTokens = 0
+			let modelTotalCost = 0
+			let modelTotalDuration = 0
+
+			for (const run of modelRuns) {
+				const tasks = db.getRunTasks(run.id)
+				modelTasks += tasks.length
+
+				for (const task of tasks) {
+					if (task.success) {
+						modelSuccessfulTasks++
+					}
+
+					const metrics = db.getTaskMetrics(task.id)
+
+					const tokensIn = metrics.find((m) => m.name === "tokensIn")?.value || 0
+					const tokensOut = metrics.find((m) => m.name === "tokensOut")?.value || 0
+					modelTotalTokens += tokensIn + tokensOut
+
+					modelTotalCost += metrics.find((m) => m.name === "cost")?.value || 0
+					modelTotalDuration += metrics.find((m) => m.name === "duration")?.value || 0
+				}
+			}
+
+			modelSummary.tasks = modelTasks
+			modelSummary.successRate = modelTasks > 0 ? modelSuccessfulTasks / modelTasks : 0
+			modelSummary.averageTokens = modelTasks > 0 ? modelTotalTokens / modelTasks : 0
+			modelSummary.averageCost = modelTasks > 0 ? modelTotalCost / modelTasks : 0
+			modelSummary.averageDuration = modelTasks > 0 ? modelTotalDuration / modelTasks : 0
+
+			modelReports[model] = modelSummary
+		}
+
+		// Save reports
+		const reportDir = path.join(path.resolve(__dirname, "../../../"), "results", "reports")
+		fs.mkdirSync(reportDir, { recursive: true })
+
+		const timestamp = new Date().toISOString().replace(/:/g, "-")
+
+		if (format === "json") {
+			// Save JSON reports
+			fs.writeFileSync(path.join(reportDir, `summary-${timestamp}.json`), JSON.stringify(summary, null, 2))
+
+			fs.writeFileSync(path.join(reportDir, `benchmarks-${timestamp}.json`), JSON.stringify(benchmarkReports, null, 2))
+
+			fs.writeFileSync(path.join(reportDir, `models-${timestamp}.json`), JSON.stringify(modelReports, null, 2))
+
+			spinner.succeed(`JSON reports generated in ${reportDir}`)
+		} else {
+			// Generate markdown report
+			const outputPath = options.output || path.join(reportDir, `report-${timestamp}.md`)
+
+			generateMarkdownReport(summary, benchmarkReports, modelReports, outputPath)
+
+			spinner.succeed(`Markdown report generated at ${outputPath}`)
+		}
+	} catch (error: any) {
+		console.error(chalk.red(`Error generating report: ${error.message}`))
+		console.error(error.stack)
+	} finally {
+		db.close()
+	}
+}

evals/cli/src/commands/run.ts

@@ -0,0 +1,133 @@
+import * as path from "path"
+import { v4 as uuidv4 } from "uuid"
+import chalk from "chalk"
+import ora from "ora"
+import { getAdapter } from "../adapters"
+import { ResultsDatabase } from "../db"
+import { spawnVSCode, cleanupVSCode } from "../utils/vscode"
+import { sendTaskToServer } from "../utils/task"
+import { storeTaskResult } from "../utils/results"
+
+interface RunOptions {
+	benchmark?: string
+	model: string
+	count?: number
+	apiKey?: string
+}
+
+/**
+ * Handler for the run command
+ * @param options Command options
+ */
+export async function runHandler(options: RunOptions): Promise<void> {
+	// Determine which benchmarks to run
+	const benchmarks = options.benchmark ? [options.benchmark] : ["exercism"] // Default to exercism for now
+	const model = options.model
+	const count = options.count || Infinity
+
+	console.log(chalk.blue(`Running evaluations for model: ${model}`))
+	console.log(chalk.blue(`Benchmarks: ${benchmarks.join(", ")}`))
+
+	// Create a run for each benchmark
+	for (const benchmark of benchmarks) {
+		const runId = uuidv4()
+		const db = new ResultsDatabase()
+
+		console.log(chalk.green(`\nStarting run for benchmark: ${benchmark}`))
+
+		// Create run in database
+		db.createRun(runId, model, benchmark)
+
+		// Get adapter for this benchmark
+		try {
+			const adapter = getAdapter(benchmark)
+
+			// List tasks
+			const spinner = ora("Listing tasks...").start()
+			const tasks = await adapter.listTasks()
+			spinner.succeed(`Found ${tasks.length} tasks for ${benchmark}`)
+
+			// Limit number of tasks if specified
+			const tasksToRun = tasks.slice(0, count)
+
+			console.log(chalk.blue(`Running ${tasksToRun.length} tasks...`))
+
+			// Run each task
+			for (let i = 0; i < tasksToRun.length; i++) {
+				const task = tasksToRun[i]
+
+				console.log(chalk.cyan(`\nTask ${i + 1}/${tasksToRun.length}: ${task.name}`))
+
+				// Prepare task
+				const prepareSpinner = ora("Preparing task...").start()
+				const preparedTask = await adapter.prepareTask(task.id)
+				prepareSpinner.succeed("Task prepared")
+
+				// Spawn VSCode
+				console.log("Spawning VSCode...")
+				await spawnVSCode(preparedTask.workspacePath)
+
+				// Send task to server
+				const sendSpinner = ora("Sending task to server...").start()
+				try {
+					const result = await sendTaskToServer(preparedTask.description, options.apiKey)
+					sendSpinner.succeed("Task completed")
+
+					// Verify result
+					const verifySpinner = ora("Verifying result...").start()
+					const verification = await adapter.verifyResult(preparedTask, result)
+
+					if (verification.success) {
+						verifySpinner.succeed(
+							`Verification successful: ${verification.metrics.testsPassed}/${verification.metrics.testsTotal} tests passed`,
+						)
+					} else {
+						verifySpinner.fail(
+							`Verification failed: ${verification.metrics.testsPassed}/${verification.metrics.testsTotal} tests passed`,
+						)
+					}
+
+					// Store result
+					const storeSpinner = ora("Storing result...").start()
+					await storeTaskResult(runId, preparedTask, result, verification)
+					storeSpinner.succeed("Result stored")
+
+					console.log(chalk.green(`Task completed. Success: ${verification.success}`))
+
+					// Clean up VS Code and temporary files
+					const cleanupSpinner = ora("Cleaning up...").start()
+					try {
+						await cleanupVSCode(preparedTask.workspacePath)
+						cleanupSpinner.succeed("Cleanup completed")
+					} catch (cleanupError: any) {
+						cleanupSpinner.fail(`Cleanup failed: ${cleanupError.message}`)
+						console.error(chalk.yellow(cleanupError.stack))
+					}
+				} catch (error: any) {
+					sendSpinner.fail(`Task failed: ${error.message}`)
+					console.error(chalk.red(error.stack))
+
+					// Clean up VS Code and temporary files even if the task failed
+					const cleanupSpinner = ora("Cleaning up...").start()
+					try {
+						await cleanupVSCode(preparedTask.workspacePath)
+						cleanupSpinner.succeed("Cleanup completed")
+					} catch (cleanupError: any) {
+						cleanupSpinner.fail(`Cleanup failed: ${cleanupError.message}`)
+						console.error(chalk.yellow(cleanupError.stack))
+					}
+				}
+			}
+
+			// Mark run as complete
+			db.completeRun(runId)
+
+			console.log(chalk.green(`\nRun complete for benchmark: ${benchmark}`))
+		} catch (error: any) {
+			console.error(chalk.red(`Error running benchmark ${benchmark}: ${error.message}`))
+			console.error(error.stack)
+		}
+	}
+
+	console.log(chalk.green("\nAll evaluations complete"))
+}

evals/cli/src/commands/setup.ts

@@ -0,0 +1,72 @@
+import * as path from "path"
+import * as fs from "fs"
+import execa from "execa"
+import chalk from "chalk"
+import ora from "ora"
+import { getAllAdapters } from "../adapters/index"
+import { BenchmarkAdapter } from "../adapters/types"
+
+interface SetupOptions {
+	benchmarks: string
+}
+
+/**
+ * Handler for the setup command
+ * @param options Command options
+ */
+export async function setupHandler(options: SetupOptions): Promise<void> {
+	const benchmarks = options.benchmarks.split(",")
+
+	console.log(chalk.blue(`Setting up benchmarks: ${benchmarks.join(", ")}`))
+
+	// Create directories
+	const evalsDir = path.resolve(__dirname, "../../../")
+	const reposDir = path.join(evalsDir, "repositories")
+	const resultsDir = path.join(evalsDir, "results")
+
+	const spinner = ora("Creating directory structure").start()
+
+	try {
+		fs.mkdirSync(reposDir, { recursive: true })
+		fs.mkdirSync(resultsDir, { recursive: true })
+		fs.mkdirSync(path.join(resultsDir, "runs"), { recursive: true })
+		fs.mkdirSync(path.join(resultsDir, "reports"), { recursive: true })
+		spinner.succeed("Directory structure created")
+	} catch (error) {
+		spinner.fail(`Failed to create directory structure: ${(error as Error).message}`)
+		throw error
+	}
+
+	// Set up each benchmark
+	try {
+		const adapters = getAllAdapters().filter((adapter: BenchmarkAdapter) => benchmarks.includes(adapter.name))
+
+		if (adapters.length === 0) {
+			console.warn(chalk.yellow("No valid benchmarks specified. Available benchmarks:"))
+			console.warn(
+				chalk.yellow(
+					getAllAdapters()
+						.map((a: BenchmarkAdapter) => a.name)
+						.join(", "),
+				),
+			)
+			return
+		}
+
+		for (const adapter of adapters) {
+			const setupSpinner = ora(`Setting up ${adapter.name}...`).start()
+			try {
+				await adapter.setup()
+				setupSpinner.succeed(`${adapter.name} setup complete`)
+			} catch (error) {
+				setupSpinner.fail(`Failed to set up ${adapter.name}: ${(error as Error).message}`)
+				throw error
+			}
+		}
+
+		console.log(chalk.green("Setup complete"))
+	} catch (error) {
+		console.error(chalk.red(`Setup failed: ${(error as Error).message}`))
+		throw error
+	}
+}

evals/cli/src/db/index.ts

@@ -0,0 +1,211 @@
+import * as path from "path"
+import * as fs from "fs"
+import Database from "better-sqlite3"
+import { SCHEMA } from "./schema"
+
+const EVALS_DIR = path.resolve(__dirname, "../../../")
+
+/**
+ * Database class for storing evaluation results
+ */
+export class ResultsDatabase {
+	db: Database.Database
+
+	constructor() {
+		// Ensure results directory exists
+		const resultsDir = path.join(EVALS_DIR, "results")
+		fs.mkdirSync(resultsDir, { recursive: true })
+
+		// Create database file
+		const dbPath = path.join(resultsDir, "evals.db")
+		this.db = new Database(dbPath)
+
+		// Initialize schema
+		this.initSchema()
+	}
+
+	/**
+	 * Initialize the database schema
+	 */
+	private initSchema(): void {
+		this.db.exec(SCHEMA)
+	}
+
+	/**
+	 * Create a new evaluation run
+	 * @param id Run ID
+	 * @param model Model name
+	 * @param benchmark Benchmark name
+	 */
+	createRun(id: string, model: string, benchmark: string): void {
+		const stmt = this.db.prepare(`
+      INSERT INTO runs (id, timestamp, model, benchmark)
+      VALUES (?, ?, ?, ?)
+    `)
+
+		stmt.run(id, Date.now(), model, benchmark)
+	}
+
+	/**
+	 * Mark a run as completed
+	 * @param id Run ID
+	 */
+	completeRun(id: string): void {
+		const stmt = this.db.prepare(`
+      UPDATE runs SET completed = 1 WHERE id = ?
+    `)
+
+		stmt.run(id)
+	}
+
+	/**
+	 * Create a new task
+	 * @param id Task ID
+	 * @param runId Run ID
+	 * @param taskId Original task ID
+	 */
+	createTask(id: string, runId: string, taskId: string): void {
+		const stmt = this.db.prepare(`
+      INSERT INTO tasks (id, run_id, task_id, timestamp)
+      VALUES (?, ?, ?, ?)
+    `)
+
+		stmt.run(id, runId, taskId, Date.now())
+	}
+
+	/**
+	 * Mark a task as completed
+	 * @param id Task ID
+	 * @param success Whether the task was successful
+	 * @param toolCalls Total tool calls
+	 * @param toolFailures Total tool failures
+	 */
+	completeTask(id: string, success: boolean, toolCalls: number = 0, toolFailures: number = 0): void {
+		const stmt = this.db.prepare(`
+      UPDATE tasks 
+      SET success = ?, total_tool_calls = ?, total_tool_failures = ? 
+      WHERE id = ?
+    `)
+
+		stmt.run(success ? 1 : 0, toolCalls, toolFailures, id)
+	}
+
+	/**
+	 * Add a metric to a task
+	 * @param taskId Task ID
+	 * @param name Metric name
+	 * @param value Metric value
+	 */
+	addMetric(taskId: string, name: string, value: number): void {
+		const stmt = this.db.prepare(`
+      INSERT INTO metrics (task_id, name, value)
+      VALUES (?, ?, ?)
+    `)
+
+		stmt.run(taskId, name, value)
+	}
+
+	/**
+	 * Add a tool call record
+	 * @param taskId Task ID
+	 * @param toolName Tool name
+	 * @param callCount Number of calls
+	 * @param failureCount Number of failures
+	 */
+	addToolCall(taskId: string, toolName: string, callCount: number, failureCount: number): void {
+		const stmt = this.db.prepare(`
+      INSERT INTO tool_calls (task_id, tool_name, call_count, failure_count)
+      VALUES (?, ?, ?, ?)
+    `)
+
+		stmt.run(taskId, toolName, callCount, failureCount)
+	}
+
+	/**
+	 * Add a file record
+	 * @param taskId Task ID
+	 * @param filePath File path
+	 * @param status File status (created, modified, deleted)
+	 */
+	addFile(taskId: string, filePath: string, status: "created" | "modified" | "deleted"): void {
+		const stmt = this.db.prepare(`
+      INSERT INTO files (task_id, path, status)
+      VALUES (?, ?, ?)
+    `)
+
+		stmt.run(taskId, filePath, status)
+	}
+
+	/**
+	 * Get all runs
+	 * @returns Array of runs
+	 */
+	getRuns(): any[] {
+		const stmt = this.db.prepare(`
+      SELECT * FROM runs ORDER BY timestamp DESC
+    `)
+
+		return stmt.all()
+	}
+
+	/**
+	 * Get all tasks for a run
+	 * @param runId Run ID
+	 * @returns Array of tasks
+	 */
+	getRunTasks(runId: string): any[] {
+		const stmt = this.db.prepare(`
+      SELECT * FROM tasks WHERE run_id = ? ORDER BY timestamp ASC
+    `)
+
+		return stmt.all(runId)
+	}
+
+	/**
+	 * Get all metrics for a task
+	 * @param taskId Task ID
+	 * @returns Array of metrics
+	 */
+	getTaskMetrics(taskId: string): any[] {
+		const stmt = this.db.prepare(`
+      SELECT name, value FROM metrics WHERE task_id = ?
+    `)
+
+		return stmt.all(taskId)
+	}
+
+	/**
+	 * Get all tool calls for a task
+	 * @param taskId Task ID
+	 * @returns Array of tool calls
+	 */
+	getTaskToolCalls(taskId: string): any[] {
+		const stmt = this.db.prepare(`
+      SELECT tool_name, call_count, failure_count 
+      FROM tool_calls 
+      WHERE task_id = ?
+    `)
+
+		return stmt.all(taskId)
+	}
+
+	/**
+	 * Get all files for a task
+	 * @param taskId Task ID
+	 * @returns Array of files
+	 */
+	getTaskFiles(taskId: string): any[] {
+		const stmt = this.db.prepare(`
+      SELECT path, status FROM files WHERE task_id = ?
+    `)
+
+		return stmt.all(taskId)
+	}
+
+	/**
+	 * Close the database connection
+	 */
+	close(): void {
+		this.db.close()
+	}
+}

evals/cli/src/db/schema.ts

@@ -0,0 +1,48 @@
+/**
+ * SQL schema for the evaluation database
+ */
+export const SCHEMA = `
+CREATE TABLE IF NOT EXISTS runs (
+  id TEXT PRIMARY KEY,
+  timestamp INTEGER NOT NULL,
+  model TEXT NOT NULL,
+  benchmark TEXT NOT NULL,
+  completed INTEGER NOT NULL DEFAULT 0
+);
+
+CREATE TABLE IF NOT EXISTS tasks (
+  id TEXT PRIMARY KEY,
+  run_id TEXT NOT NULL,
+  task_id TEXT NOT NULL,
+  timestamp INTEGER NOT NULL,
+  success INTEGER NOT NULL DEFAULT 0,
+  total_tool_calls INTEGER DEFAULT 0,
+  total_tool_failures INTEGER DEFAULT 0,
+  FOREIGN KEY (run_id) REFERENCES runs(id)
+);
+
+CREATE TABLE IF NOT EXISTS metrics (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  task_id TEXT NOT NULL,
+  name TEXT NOT NULL,
+  value REAL NOT NULL,
+  FOREIGN KEY (task_id) REFERENCES tasks(id)
+);
+
+CREATE TABLE IF NOT EXISTS tool_calls (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  task_id TEXT NOT NULL,
+  tool_name TEXT NOT NULL,
+  call_count INTEGER NOT NULL,
+  failure_count INTEGER NOT NULL,
+  FOREIGN KEY (task_id) REFERENCES tasks(id)
+);
+
+CREATE TABLE IF NOT EXISTS files (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  task_id TEXT NOT NULL,
+  path TEXT NOT NULL,
+  status TEXT NOT NULL,
+  FOREIGN KEY (task_id) REFERENCES tasks(id)
+);
+`

evals/cli/src/index.ts

@@ -0,0 +1,86 @@
+#!/usr/bin/env node
+import { Command } from "commander"
+import chalk from "chalk"
+import { setupHandler } from "./commands/setup"
+import { runHandler } from "./commands/run"
+import { reportHandler } from "./commands/report"
+import { evalsEnvHandler } from "./commands/evals-env"
+
+// Create the CLI program
+const program = new Command()
+
+// Set up CLI metadata
+program.name("cline-eval").description("CLI tool for orchestrating Cline evaluations across multiple benchmarks").version("0.1.0")
+
+// Setup command
+program
+	.command("setup")
+	.description("Clone and set up benchmark repositories")
+	.option(
+		"-b, --benchmarks <benchmarks>",
+		"Comma-separated list of benchmarks to set up",
+		"exercism,swe-bench,swelancer,multi-swe",
+	)
+	.action(async (options) => {
+		try {
+			await setupHandler(options)
+		} catch (error) {
+			console.error(chalk.red(`Error during setup: ${error instanceof Error ? error.message : String(error)}`))
+			process.exit(1)
+		}
+	})
+
+// Run command
+program
+	.command("run")
+	.description("Run evaluations")
+	.option("-b, --benchmark <benchmark>", "Specific benchmark to run")
+	.option("-m, --model <model>", "Model to evaluate", "claude-3-opus-20240229")
+	.option("-c, --count <count>", "Number of tasks to run", parseInt)
+	.option("-k, --api-key <apiKey>", "Cline API key to use for evaluations")
+	.action(async (options) => {
+		try {
+			await runHandler(options)
+		} catch (error) {
+			console.error(chalk.red(`Error during run: ${error instanceof Error ? error.message : String(error)}`))
+			process.exit(1)
+		}
+	})
+
+// Report command
+program
+	.command("report")
+	.description("Generate reports")
+	.option("-f, --format <format>", "Report format (json, markdown)", "markdown")
+	.option("-o, --output <path>", "Output path for the report")
+	.action(async (options) => {
+		try {
+			await reportHandler(options)
+		} catch (error) {
+			console.error(chalk.red(`Error generating report: ${error instanceof Error ? error.message : String(error)}`))
+			process.exit(1)
+		}
+	})
+
+// Evals-env command
+program
+	.command("evals-env")
+	.description("Manage evals.env files for test mode activation")
+	.argument("<action>", "Action to perform: create, remove, or check")
+	.option("-d, --directory <directory>", "Directory to create/remove/check evals.env file in (defaults to current directory)")
+	.action(async (action, options) => {
+		try {
+			await evalsEnvHandler({ action, ...options })
+		} catch (error) {
+			console.error(chalk.red(`Error managing evals.env file: ${error instanceof Error ? error.message : String(error)}`))
+			process.exit(1)
+		}
+	})
+
+// Parse command line arguments
+program.parse(process.argv)
+
+// If no arguments provided, show help
+if (process.argv.length === 2) {
+	program.help()
+}

evals/cli/src/utils/evals-env.ts

@@ -0,0 +1,79 @@
+import * as fs from "fs"
+import * as path from "path"
+import chalk from "chalk"
+
+/**
+ * Creates an evals.env file in the specified directory
+ * @param directory The directory where the evals.env file should be created
+ * @returns True if the file was created, false if it already exists
+ */
+export function createEvalsEnvFile(directory: string): boolean {
+	const evalsEnvPath = path.join(directory, "evals.env")
+
+	// Check if the file already exists
+	if (fs.existsSync(evalsEnvPath)) {
+		console.log(chalk.yellow(`evals.env file already exists at ${evalsEnvPath}`))
+		return false
+	}
+
+	// Create the file
+	try {
+		const content = `# This file activates Cline test mode
+# Created at: ${new Date().toISOString()}
+# 
+# This file is automatically detected by the Cline extension
+# and enables test mode for automated evaluations.
+#
+# Delete this file to deactivate test mode.
+`
+		fs.writeFileSync(evalsEnvPath, content)
+		console.log(chalk.green(`Created evals.env file at ${evalsEnvPath}`))
+		return true
+	} catch (error) {
+		console.error(chalk.red(`Error creating evals.env file: ${error}`))
+		return false
+	}
+}
+
+/**
+ * Removes an evals.env file from the specified directory
+ * @param directory The directory where the evals.env file should be removed
+ * @returns True if the file was removed, false if it doesn't exist
+ */
+export function removeEvalsEnvFile(directory: string): boolean {
+	const evalsEnvPath = path.join(directory, "evals.env")
+
+	// Check if the file exists
+	if (!fs.existsSync(evalsEnvPath)) {
+		console.log(chalk.yellow(`No evals.env file found at ${evalsEnvPath}`))
+		return false
+	}
+
+	// Remove the file
+	try {
+		fs.unlinkSync(evalsEnvPath)
+		console.log(chalk.green(`Removed evals.env file from ${evalsEnvPath}`))
+		return true
+	} catch (error) {
+		console.error(chalk.red(`Error removing evals.env file: ${error}`))
+		return false
+	}
+}
+
+/**
+ * Checks if an evals.env file exists in the specified directory
+ * @param directory The directory to check for an evals.env file
+ * @returns True if the file exists, false otherwise
+ */
+export function checkEvalsEnvFile(directory: string): boolean {
+	const evalsEnvPath = path.join(directory, "evals.env")
+	const exists = fs.existsSync(evalsEnvPath)
+
+	if (exists) {
+		console.log(chalk.green(`evals.env file found at ${evalsEnvPath}`))
+	} else {
+		console.log(chalk.yellow(`No evals.env file found at ${evalsEnvPath}`))
+	}
+
+	return exists
+}

evals/cli/src/utils/extensions.ts

@@ -0,0 +1,131 @@
+import execa from "execa"
+import * as fs from "fs"
+import * as path from "path"
+import * as os from "os"
+
+/**
+ * List of VSCode extensions to install for evaluation environments
+ * These extensions provide language support and other useful features
+ */
+export const REQUIRED_EXTENSIONS = [
+	"golang.go", // Go language support
+	"dbaeumer.vscode-eslint", // ESLint support
+	"redhat.java", // Java support
+	"ms-python.python", // Python support
+	"rust-lang.rust-analyzer", // Rust support
+	"ms-vscode.cpptools", // C/C++ support
+]
+
+/**
+ * Install required VSCode extensions in the specified extensions directory
+ * @param extensionsDir The directory where extensions should be installed
+ * @returns Promise that resolves when all extensions are installed
+ */
+export async function installRequiredExtensions(extensionsDir: string): Promise<void> {
+	console.log("Installing required VSCode extensions...")
+
+	// Create the extensions directory if it doesn't exist
+	if (!fs.existsSync(extensionsDir)) {
+		fs.mkdirSync(extensionsDir, { recursive: true })
+	}
+
+	// Install each extension
+	for (const extension of REQUIRED_EXTENSIONS) {
+		try {
+			console.log(`Installing extension: ${extension}...`)
+			await execa("code", ["--extensions-dir", extensionsDir, "--install-extension", extension, "--force"])
+			console.log(`✅ Extension ${extension} installed successfully`)
+		} catch (error: any) {
+			console.warn(`⚠️ Failed to install extension ${extension}: ${error.message}`)
+			// Continue with other extensions even if one fails
+		}
+	}
+
+	console.log("✅ All required extensions installed")
+}
+
+/**
+ * Check if a VSCode extension is installed in the specified directory
+ * @param extensionsDir The directory to check for installed extensions
+ * @param extensionId The ID of the extension to check
+ * @returns True if the extension is installed, false otherwise
+ */
+export function isExtensionInstalled(extensionsDir: string, extensionId: string): boolean {
+	// Extensions are installed in directories named publisher.name-version
+	// We need to check if any directory starts with the extensionId
+	const extensionPrefix = extensionId.toLowerCase() + "-"
+
+	try {
+		const files = fs.readdirSync(extensionsDir)
+		return files.some((file) => {
+			const lowerCaseFile = file.toLowerCase()
+			return lowerCaseFile === extensionId.toLowerCase() || lowerCaseFile.startsWith(extensionPrefix)
+		})
+	} catch (error) {
+		return false
+	}
+}
+
+/**
+ * Get the path to the VSCode settings file in the specified user data directory
+ * @param userDataDir The VSCode user data directory
+ * @returns The path to the settings.json file
+ */
+export function getSettingsPath(userDataDir: string): string {
+	const settingsDir = path.join(userDataDir, "User")
+	fs.mkdirSync(settingsDir, { recursive: true })
+	return path.join(settingsDir, "settings.json")
+}
+
+/**
+ * Configure extension settings in the VSCode user data directory
+ * @param userDataDir The VSCode user data directory
+ */
+export function configureExtensionSettings(userDataDir: string): void {
+	const settingsPath = getSettingsPath(userDataDir)
+
+	// Read existing settings if they exist
+	let settings = {}
+	if (fs.existsSync(settingsPath)) {
+		try {
+			settings = JSON.parse(fs.readFileSync(settingsPath, "utf8"))
+		} catch (error) {
+			console.warn(`Error reading settings file: ${error}`)
+		}
+	}
+
+	// Add or update extension-specific settings
+	const updatedSettings = {
+		...settings,
+		// Go extension settings
+		"go.toolsManagement.autoUpdate": false,
+		"go.survey.prompt": false,
+
+		// ESLint settings
+		"eslint.enable": true,
+		"eslint.run": "onSave",
+
+		// Java settings
+		"java.configuration.checkProjectSettingsExclusions": false,
+		"java.configure.checkForOutdatedExtensions": false,
+		"java.help.firstView": false,
+
+		// Python settings
+		"python.experiments.enabled": false,
+		"python.showStartPage": false,
+
+		// Rust settings
+		"rust-analyzer.checkOnSave.command": "check",
+
+		// C/C++ settings
+		"C_Cpp.intelliSenseEngine": "default",
+
+		// General extension settings
+		"extensions.autoUpdate": false,
+		"extensions.ignoreRecommendations": true,
+	}
+
+	// Write updated settings
+	fs.writeFileSync(settingsPath, JSON.stringify(updatedSettings, null, 2))
+	console.log("✅ Extension settings configured")
+}

evals/cli/src/utils/markdown.ts

@@ -0,0 +1,109 @@
+import * as fs from "fs"
+import * as path from "path"
+
+/**
+ * Generate a markdown report from evaluation results
+ * @param summary Overall summary
+ * @param benchmarkReports Benchmark-specific reports
+ * @param modelReports Model-specific reports
+ * @param outputPath Output file path
+ */
+export function generateMarkdownReport(
+	summary: any,
+	benchmarkReports: Record<string, any>,
+	modelReports: Record<string, any>,
+	outputPath: string,
+): void {
+	let markdown = `# Cline Evaluation Report\n\n`
+
+	// Generate summary section
+	markdown += `## Summary\n\n`
+	markdown += `- **Total Runs:** ${summary.runs}\n`
+	markdown += `- **Models:** ${summary.models.join(", ")}\n`
+	markdown += `- **Benchmarks:** ${summary.benchmarks.join(", ")}\n`
+	markdown += `- **Total Tasks:** ${summary.tasks}\n`
+	markdown += `- **Success Rate:** ${(summary.successRate * 100).toFixed(2)}%\n`
+	markdown += `- **Average Tokens:** ${Math.round(summary.averageTokens)}\n`
+	markdown += `- **Average Cost:** $${summary.averageCost.toFixed(4)}\n`
+	markdown += `- **Average Duration:** ${(summary.averageDuration / 1000).toFixed(2)}s\n`
+	markdown += `- **Total Tool Calls:** ${summary.totalToolCalls}\n`
+	markdown += `- **Tool Success Rate:** ${(summary.toolSuccessRate * 100).toFixed(2)}%\n\n`
+
+	// Generate tool usage section
+	markdown += `## Tool Usage\n\n`
+	markdown += `| Tool | Calls | Failures | Success Rate |\n`
+	markdown += `| ---- | ----- | -------- | ------------ |\n`
+
+	for (const [toolName, metrics] of Object.entries(summary.toolUsage)) {
+		const calls = (metrics as any).calls
+		const failures = (metrics as any).failures
+		const successRate = calls > 0 ? (1 - failures / calls) * 100 : 100
+
+		markdown += `| ${toolName} | ${calls} | ${failures} | ${successRate.toFixed(2)}% |\n`
+	}
+
+	// Generate benchmark results section
+	markdown += `\n## Benchmark Results\n\n`
+
+	for (const [benchmark, report] of Object.entries(benchmarkReports)) {
+		markdown += `### ${benchmark}\n\n`
+		markdown += `- **Runs:** ${report.runs}\n`
+		markdown += `- **Models:** ${report.models.join(", ")}\n`
+		markdown += `- **Tasks:** ${report.tasks}\n`
+		markdown += `- **Success Rate:** ${(report.successRate * 100).toFixed(2)}%\n`
+		markdown += `- **Average Tokens:** ${Math.round(report.averageTokens)}\n`
+		markdown += `- **Average Cost:** $${report.averageCost.toFixed(4)}\n`
+		markdown += `- **Average Duration:** ${(report.averageDuration / 1000).toFixed(2)}s\n\n`
+	}
+
+	// Generate model results section
+	markdown += `## Model Results\n\n`
+
+	for (const [model, report] of Object.entries(modelReports)) {
+		markdown += `### ${model}\n\n`
+		markdown += `- **Runs:** ${report.runs}\n`
+		markdown += `- **Benchmarks:** ${report.benchmarks.join(", ")}\n`
+		markdown += `- **Tasks:** ${report.tasks}\n`
+		markdown += `- **Success Rate:** ${(report.successRate * 100).toFixed(2)}%\n`
+		markdown += `- **Average Tokens:** ${Math.round(report.averageTokens)}\n`
+		markdown += `- **Average Cost:** $${report.averageCost.toFixed(4)}\n`
+		markdown += `- **Average Duration:** ${(report.averageDuration / 1000).toFixed(2)}s\n\n`
+	}
+
+	// Add charts using Mermaid
+	markdown += `## Charts\n\n`
+
+	// Success rate by benchmark chart
+	markdown += `### Success Rate by Benchmark\n\n`
+	markdown += "```mermaid\n"
+	markdown += "graph TD\n"
+	markdown += "  title[Success Rate by Benchmark]\n"
+	markdown += "  style title fill:none,stroke:none\n\n"
+
+	for (const [benchmark, report] of Object.entries(benchmarkReports)) {
+		const successRate = (report.successRate * 100).toFixed(2)
+		markdown += `  ${benchmark}[${benchmark}: ${successRate}%]\n`
+	}
+
+	markdown += "```\n\n"
+
+	// Success rate by model chart
+	markdown += `### Success Rate by Model\n\n`
+	markdown += "```mermaid\n"
+	markdown += "graph TD\n"
+	markdown += "  title[Success Rate by Model]\n"
+	markdown += "  style title fill:none,stroke:none\n\n"
+
+	for (const [model, report] of Object.entries(modelReports)) {
+		const successRate = (report.successRate * 100).toFixed(2)
+		markdown += `  ${model.replace(/[-\.]/g, "_")}[${model}: ${successRate}%]\n`
+	}
+
+	markdown += "```\n\n"
+
+	// Add timestamp
+	markdown += `\n\n---\n\nReport generated on ${new Date().toISOString()}\n`
+
+	// Write markdown to file
+	fs.writeFileSync(outputPath, markdown)
+}

evals/cli/src/utils/results.ts

@@ -0,0 +1,79 @@
+import { v4 as uuidv4 } from "uuid"
+import { ResultsDatabase } from "../db"
+import { Task } from "../adapters/types"
+
+/**
+ * Store task result in the database
+ * @param runId The run ID
+ * @param task The task that was executed
+ * @param result The result from the test server
+ * @param verification The verification result
+ */
+export async function storeTaskResult(runId: string, task: Task, result: any, verification: any): Promise<void> {
+	const db = new ResultsDatabase()
+	const taskId = uuidv4()
+
+	try {
+		// Extract metrics from the result
+		const { metrics } = result
+		const totalToolCalls = metrics?.totalToolCalls || 0
+		const totalToolFailures = metrics?.totalToolFailures || 0
+
+		// Create task with tool metrics
+		db.createTask(taskId, runId, task.id)
+		db.completeTask(taskId, verification.success, totalToolCalls, totalToolFailures)
+
+		// Store metrics
+		if (metrics) {
+			// Store token metrics
+			if (metrics.tokensIn) db.addMetric(taskId, "tokensIn", metrics.tokensIn)
+			if (metrics.tokensOut) db.addMetric(taskId, "tokensOut", metrics.tokensOut)
+			if (metrics.cost) db.addMetric(taskId, "cost", metrics.cost)
+			if (metrics.duration) db.addMetric(taskId, "duration", metrics.duration)
+
+			// Store tool call metrics
+			if (metrics.toolCalls) {
+				for (const [toolName, callCount] of Object.entries(metrics.toolCalls)) {
+					const failureCount = metrics.toolFailures?.[toolName] || 0
+					db.addToolCall(taskId, toolName, callCount as number, failureCount)
+				}
+			}
+		}
+
+		// Store verification metrics
+		if (verification.metrics) {
+			for (const [key, value] of Object.entries(verification.metrics)) {
+				if (typeof value === "number") {
+					db.addMetric(taskId, key, value)
+				}
+			}
+		}
+
+		// Store file changes
+		if (result.files) {
+			// Store created files
+			if (result.files.created) {
+				for (const file of result.files.created) {
+					db.addFile(taskId, file, "created")
+				}
+			}
+
+			// Store modified files
+			if (result.files.modified) {
+				for (const file of result.files.modified) {
+					db.addFile(taskId, file, "modified")
+				}
+			}
+
+			// Store deleted files
+			if (result.files.deleted) {
+				for (const file of result.files.deleted) {
+					db.addFile(taskId, file, "deleted")
+				}
+			}
+		}
+	} finally {
+		// Close the database connection
+		db.close()
+	}
+}

evals/cli/src/utils/task.ts

@@ -0,0 +1,52 @@
+import fetch from "node-fetch"
+import chalk from "chalk"
+
+/**
+ * Send a task to the Cline test server
+ * @param task The task description to send
+ * @param apiKey Optional Cline API key to use for the task
+ * @returns The result of the task execution
+ */
+export async function sendTaskToServer(task: string, apiKey?: string): Promise<any> {
+	const SERVER_URL = "http://localhost:9876/task"
+
+	try {
+		console.log(chalk.blue(`Sending task to server: ${task.substring(0, 100)}${task.length > 100 ? "..." : ""}`))
+
+		const response = await fetch(SERVER_URL, {
+			method: "POST",
+			headers: {
+				"Content-Type": "application/json",
+			},
+			body: JSON.stringify({
+				task,
+				apiKey,
+			}),
+		})
+
+		if (!response.ok) {
+			const errorText = await response.text()
+			throw new Error(`Server responded with status ${response.status}: ${errorText}`)
+		}
+
+		const result = await response.json()
+
+		if (!result.success) {
+			throw new Error(`Task execution failed: ${result.error || "Unknown error"}`)
+		}
+
+		if (result.timeout) {
+			throw new Error("Task execution timed out")
+		}
+
+		return result
+	} catch (error: any) {
+		if (error.code === "ECONNREFUSED") {
+			throw new Error(
+				"Could not connect to the test server. Make sure VSCode is running with the Cline extension and the test server is active.",
+			)
+		}
+
+		throw error
+	}
+}

evals/cli/src/utils/vscode.ts

@@ -0,0 +1,615 @@
+import execa from "execa"
+import * as path from "path"
+import * as fs from "fs"
+import fetch from "node-fetch"
+import * as os from "os"
+import { installRequiredExtensions, configureExtensionSettings } from "./extensions"
+
+// Store temporary directories for cleanup
+interface VSCodeResources {
+	tempUserDataDir: string
+	tempExtensionsDir: string
+	vscodePid?: number
+}
+
+// Global map to track resources for each workspace
+const workspaceResources = new Map<string, VSCodeResources>()
+
+/**
+ * Spawn a VSCode instance with the Cline extension
+ * @param workspacePath The workspace path to open
+ * @param vsixPath Optional path to a VSIX file to install
+ * @returns The resources created for this VS Code instance
+ */
+export async function spawnVSCode(workspacePath: string, vsixPath?: string): Promise<VSCodeResources> {
+	// Ensure the workspace path exists
+	if (!fs.existsSync(workspacePath)) {
+		throw new Error(`Workspace path does not exist: ${workspacePath}`)
+	}
+
+	// If no VSIX path is provided, build one with IS_TEST=true
+	if (!vsixPath) {
+		try {
+			// Build the VSIX (no longer need to set IS_TEST=true as we'll use evals.env file)
+			console.log("Building VSIX...")
+			const clineRoot = path.resolve(process.cwd(), "..", "..")
+			await execa("npx", ["vsce", "package"], {
+				cwd: clineRoot,
+				stdio: "inherit",
+			})
+
+			// Find the generated VSIX file(s)
+			const files = fs.readdirSync(clineRoot)
+			const vsixFiles = files.filter((file) => file.endsWith(".vsix"))
+
+			if (vsixFiles.length > 0) {
+				// Get file stats to find the most recent one
+				const vsixFilesWithStats = vsixFiles.map((file) => {
+					const filePath = path.join(clineRoot, file)
+					return {
+						file,
+						path: filePath,
+						mtime: fs.statSync(filePath).mtime,
+					}
+				})
+
+				// Sort by modification time (most recent first)
+				vsixFilesWithStats.sort((a, b) => b.mtime.getTime() - a.mtime.getTime())
+
+				// Use the most recent VSIX
+				vsixPath = vsixFilesWithStats[0].path
+				console.log(`Using most recent VSIX: ${vsixPath} (modified ${vsixFilesWithStats[0].mtime.toISOString()})`)
+
+				// Log all found VSIX files for debugging
+				if (vsixFiles.length > 1) {
+					console.log(`Found ${vsixFiles.length} VSIX files:`)
+					vsixFilesWithStats.forEach((f) => {
+						console.log(`  - ${f.file} (modified ${f.mtime.toISOString()})`)
+					})
+				}
+			} else {
+				console.warn("Could not find generated VSIX file")
+			}
+		} catch (error) {
+			console.warn("Failed to build test VSIX:", error)
+		}
+	}
+
+	// Create a temporary user data directory for this VS Code instance
+	const tempUserDataDir = path.join(os.tmpdir(), `vscode-cline-eval-${Date.now()}`)
+	fs.mkdirSync(tempUserDataDir, { recursive: true })
+	console.log(`Created temporary user data directory: ${tempUserDataDir}`)
+
+	// Create a temporary extensions directory to ensure no other extensions are loaded
+	const tempExtensionsDir = path.join(os.tmpdir(), `vscode-cline-eval-ext-${Date.now()}`)
+	fs.mkdirSync(tempExtensionsDir, { recursive: true })
+	console.log(`Created temporary extensions directory: ${tempExtensionsDir}`)
+
+	// Create evals.env file in the workspace to trigger test mode
+	console.log(`Creating evals.env file in workspace: ${workspacePath}`)
+	const evalsEnvPath = path.join(workspacePath, "evals.env")
+	fs.writeFileSync(
+		evalsEnvPath,
+		`# This file activates Cline test mode
+# Created at: ${new Date().toISOString()}
+# 
+# This file is automatically detected by the Cline extension
+# and enables test mode for automated evaluations.
+#
+# Delete this file to deactivate test mode.
+`,
+	)
+
+	// Create settings.json in the temporary user data directory to disable workspace trust
+	// and configure Cline to auto-open on startup
+	const settingsDir = path.join(tempUserDataDir, "User")
+	fs.mkdirSync(settingsDir, { recursive: true })
+	const settingsPath = path.join(settingsDir, "settings.json")
+	const settings = {
+		// Disable workspace trust
+		"security.workspace.trust.enabled": false,
+		"security.workspace.trust.startupPrompt": "never",
+		"security.workspace.trust.banner": "never",
+		"security.workspace.trust.emptyWindow": true,
+
+		// Configure startup behavior
+		"workbench.startupEditor": "none",
+
+		// Auto-open Cline on startup
+		"cline.autoOpenOnStartup": true,
+
+		// Show the activity bar and sidebar
+		"workbench.activityBar.visible": true,
+		"workbench.sideBar.visible": true,
+		"workbench.view.extension.saoudrizwan.claude-dev-ActivityBar.visible": true,
+		"workbench.view.alwaysShowHeaderActions": true,
+		"workbench.editor.openSideBySideDirection": "right",
+
+		// Disable GitLens from opening automatically
+		"gitlens.views.repositories.autoReveal": false,
+		"gitlens.views.fileHistory.autoReveal": false,
+		"gitlens.views.lineHistory.autoReveal": false,
+		"gitlens.views.compare.autoReveal": false,
+		"gitlens.views.search.autoReveal": false,
+		"gitlens.showWelcomeOnInstall": false,
+		"gitlens.showWhatsNewAfterUpgrades": false,
+
+		// Disable other extensions that might compete for startup focus
+		"extensions.autoUpdate": false,
+	}
+	fs.writeFileSync(settingsPath, JSON.stringify(settings, null, 2))
+	console.log(`Created settings.json to disable workspace trust and auto-open Cline`)
+
+	// Create keybindings.json to automatically open Cline on startup
+	const keybindingsPath = path.join(settingsDir, "keybindings.json")
+	const keybindings = [
+		{
+			key: "alt+c",
+			command: "workbench.view.extension.saoudrizwan.claude-dev-ActivityBar",
+			when: "viewContainer.workbench.view.extension.saoudrizwan.claude-dev-ActivityBar.enabled",
+		},
+		{
+			key: "alt+shift+c",
+			command: "cline.openInNewTab",
+			when: "viewContainer.workbench.view.extension.saoudrizwan.claude-dev-ActivityBar.enabled",
+		},
+	]
+	fs.writeFileSync(keybindingsPath, JSON.stringify(keybindings, null, 2))
+	console.log(`Created keybindings.json to help with Cline activation`)
+
+	// Build the command arguments with custom user data directory
+	const args = [
+		// Use a custom user data directory to isolate this instance
+		"--user-data-dir",
+		tempUserDataDir,
+		// Use a custom extensions directory to ensure only our extension is loaded
+		"--extensions-dir",
+		tempExtensionsDir,
+		// Disable workspace trust
+		"--disable-workspace-trust",
+		"-n",
+		workspacePath,
+		// Force the extension to be activated on startup
+		"--start-up-extension",
+		"saoudrizwan.claude-dev",
+		// Run a command on startup to open Cline
+		"--command",
+		"workbench.view.extension.saoudrizwan.claude-dev-ActivityBar",
+		// Additional flags to help with extension activation
+		"--disable-gpu=false",
+		"--max-memory=4096",
+	]
+
+	// Create a startup script to run commands after VS Code launches
+	const startupScriptPath = path.join(settingsDir, "startup.js")
+	const startupScript = `
+		// This script will be executed when VS Code starts
+		setTimeout(() => {
+			// Try to open Cline in the sidebar
+			require('vscode').commands.executeCommand('workbench.view.extension.saoudrizwan.claude-dev-ActivityBar');
+			
+			// Also try to open Cline in a tab as a fallback
+			setTimeout(() => {
+				require('vscode').commands.executeCommand('cline.openInNewTab');
+			}, 5000);
+		}, 5000);
+	`
+	fs.writeFileSync(startupScriptPath, startupScript)
+	console.log(`Created startup script to activate Cline`)
+
+	// If a VSIX is provided, install it
+	if (vsixPath) {
+		if (!fs.existsSync(vsixPath)) {
+			throw new Error(`VSIX file does not exist: ${vsixPath}`)
+		}
+		args.unshift("--install-extension", vsixPath)
+	}
+
+	// Install required extensions
+	console.log("Installing required VSCode extensions...")
+	await installRequiredExtensions(tempExtensionsDir)
+
+	// Configure extension settings
+	console.log("Configuring extension settings...")
+	configureExtensionSettings(tempUserDataDir)
+
+	// Execute the command
+	try {
+		// We don't need to install extensions globally anymore since we're using a custom user data directory
+		// The VSIX will be installed in the isolated environment if provided in the args
+
+		// Launch VS Code
+		console.log("Launching VS Code...")
+		await execa("code", args, {
+			stdio: "inherit",
+		})
+
+		// Wait longer for VSCode to initialize and extension to load
+		console.log("Waiting for VS Code to initialize...")
+		await new Promise((resolve) => setTimeout(resolve, 30000))
+
+		// Create a JavaScript file that will be loaded as a VS Code extension
+		const extensionDir = path.join(tempExtensionsDir, "cline-activator")
+		fs.mkdirSync(extensionDir, { recursive: true })
+
+		// Create package.json for the extension
+		const packageJsonPath = path.join(extensionDir, "package.json")
+		const packageJson = {
+			name: "cline-activator",
+			displayName: "Cline Activator",
+			description: "Activates Cline and starts the test server",
+			version: "0.0.1",
+			engines: {
+				vscode: "^1.60.0",
+			},
+			main: "./extension.js",
+			activationEvents: ["*"],
+			contributes: {
+				commands: [
+					{
+						command: "cline-activator.activate",
+						title: "Activate Cline",
+					},
+				],
+			},
+		}
+		fs.writeFileSync(packageJsonPath, JSON.stringify(packageJson, null, 2))
+
+		// Create extension.js
+		const extensionJsPath = path.join(extensionDir, "extension.js")
+		const extensionJs = `
+			const vscode = require('vscode');
+			
+			/**
+			 * @param {vscode.ExtensionContext} context
+			 */
+			function activate(context) {
+				console.log('Cline Activator is now active!');
+				
+				// Register the command to activate Cline
+				let disposable = vscode.commands.registerCommand('cline-activator.activate', async function () {
+					try {
+						// Make sure the Cline extension is activated
+						const extension = vscode.extensions.getExtension('saoudrizwan.claude-dev');
+						if (!extension) {
+							console.error('Cline extension not found');
+							return;
+						}
+						
+						if (!extension.isActive) {
+							console.log('Activating Cline extension...');
+							await extension.activate();
+						}
+						
+						// Show the Cline sidebar
+						console.log('Opening Cline sidebar...');
+						await vscode.commands.executeCommand('workbench.view.extension.saoudrizwan.claude-dev-ActivityBar');
+						
+						// Wait a moment for the sidebar to initialize
+						await new Promise(resolve => setTimeout(resolve, 2000));
+						
+						// Also open Cline in a tab as a fallback
+						console.log('Opening Cline in a tab...');
+						await vscode.commands.executeCommand('cline.openInNewTab');
+						
+						// Wait a moment for the tab to initialize
+						await new Promise(resolve => setTimeout(resolve, 2000));
+						
+						// Create the test server if it doesn't exist
+						console.log('Creating test server...');
+						
+						// Get the visible webview instance
+						const clineRootPath = '${path.resolve(process.cwd(), "..", "..")}';
+						const visibleWebview = require(path.join(clineRootPath, 'src', 'core', 'webview')).WebviewProvider.getVisibleInstance();
+						if (visibleWebview) {
+							require(path.join(clineRootPath, 'src', 'services', 'test', 'TestServer')).createTestServer(visibleWebview);
+							console.log('Test server created successfully');
+						} else {
+							console.error('No visible webview instance found');
+						}
+					} catch (error) {
+						console.error('Error activating Cline:', error);
+					}
+				});
+				
+				context.subscriptions.push(disposable);
+				
+				// Automatically run the command after a delay
+				setTimeout(() => {
+					vscode.commands.executeCommand('cline-activator.activate');
+				}, 5000);
+			}
+			
+			function deactivate() {}
+			
+			module.exports = {
+				activate,
+				deactivate
+			}
+		`
+		fs.writeFileSync(extensionJsPath, extensionJs)
+		console.log(`Created Cline Activator extension`)
+
+		// Try multiple approaches to activate the extension
+		let serverStarted = false
+
+		// Create an activation script to run in VS Code
+		const activationScriptPath = path.join(settingsDir, "activate-cline.js")
+		const activationScript = `
+			// This script will be executed to activate Cline and start the test server
+			const vscode = require('vscode');
+			
+			// Execute the cline-activator.activate command
+			vscode.commands.executeCommand('cline-activator.activate');
+		`
+		fs.writeFileSync(activationScriptPath, activationScript)
+		console.log(`Created activation script to run in VS Code`)
+
+		// Execute the activation script
+		try {
+			console.log("Executing activation script to start Cline and test server...")
+			await execa(
+				"code",
+				[
+					"--user-data-dir",
+					tempUserDataDir,
+					"--extensions-dir",
+					tempExtensionsDir,
+					"--folder-uri",
+					`file://${workspacePath}`,
+					"--execute",
+					activationScriptPath,
+				],
+				{
+					stdio: "inherit",
+				},
+			)
+
+			// Wait for the test server to start
+			console.log("Waiting for test server to start...")
+			for (let i = 0; i < 30; i++) {
+				try {
+					// Try to connect to the test server
+					const response = await fetch("http://localhost:9876/task", {
+						method: "OPTIONS",
+						headers: {
+							"Content-Type": "application/json",
+						},
+					})
+
+					if (response.status === 204) {
+						console.log("Test server is running!")
+						serverStarted = true
+						break
+					}
+				} catch (error) {
+					// Server not started yet, wait and try again
+					await new Promise((resolve) => setTimeout(resolve, 1000))
+				}
+			}
+		} catch (error) {
+			console.warn("Failed to execute activation script:", error)
+		}
+
+		if (!serverStarted) {
+			console.warn("Test server did not start after multiple attempts")
+			console.log("You may need to manually open the Cline extension in VS Code")
+		}
+
+		// Store the resources for this workspace
+		const resources: VSCodeResources = {
+			tempUserDataDir,
+			tempExtensionsDir,
+		}
+
+		// Store in the global map
+		workspaceResources.set(workspacePath, resources)
+
+		// Return the resources
+		return resources
+	} catch (error: any) {
+		throw new Error(`Failed to spawn VSCode: ${error.message}`)
+	}
+}
+
+/**
+ * Clean up VS Code resources and shut down the test server
+ * @param workspacePath The workspace path to clean up resources for
+ */
+export async function cleanupVSCode(workspacePath: string): Promise<void> {
+	console.log(`Cleaning up VS Code resources for workspace: ${workspacePath}`)
+
+	// Get the resources for this workspace
+	const resources = workspaceResources.get(workspacePath)
+	if (!resources) {
+		console.log(`No resources found for workspace: ${workspacePath}`)
+		return
+	}
+
+	// Try to shut down the test server
+	try {
+		console.log("Shutting down test server...")
+		await fetch("http://localhost:9876/shutdown", {
+			method: "POST",
+			headers: {
+				"Content-Type": "application/json",
+			},
+		}).catch(() => {
+			// Ignore errors, the server might already be down
+		})
+	} catch (error) {
+		console.warn(`Error shutting down test server: ${error}`)
+	}
+
+	// Try to gracefully close VS Code instead of killing it
+	try {
+		console.log("Attempting to gracefully close VS Code...")
+
+		// Create a settings file that will disable the crash reporter and the exit confirmation dialog
+		const settingsDir = path.join(resources.tempUserDataDir, "User")
+		const settingsPath = path.join(settingsDir, "settings.json")
+
+		// Read existing settings if they exist
+		let settings = {}
+		if (fs.existsSync(settingsPath)) {
+			try {
+				settings = JSON.parse(fs.readFileSync(settingsPath, "utf8"))
+			} catch (error) {
+				console.warn(`Error reading settings file: ${error}`)
+			}
+		}
+
+		// Update settings to disable crash reporter and exit confirmation
+		settings = {
+			...settings,
+			"window.confirmBeforeClose": "never",
+			"telemetry.enableCrashReporter": false,
+			"window.restoreWindows": "none",
+			"window.newWindowDimensions": "default",
+		}
+
+		// Write updated settings
+		fs.writeFileSync(settingsPath, JSON.stringify(settings, null, 2))
+
+		// On macOS, use AppleScript to quit VS Code gracefully
+		if (process.platform === "darwin") {
+			try {
+				// First try AppleScript to quit VS Code gracefully
+				await execa("osascript", ["-e", 'tell application "Visual Studio Code" to quit'])
+
+				// Wait a moment for VS Code to close
+				await new Promise((resolve) => setTimeout(resolve, 2000))
+			} catch (appleScriptError) {
+				console.warn(`Error using AppleScript to quit VS Code: ${appleScriptError}`)
+			}
+		} else if (process.platform === "win32") {
+			// On Windows, try to use taskkill without /F first
+			try {
+				await execa("taskkill", ["/IM", "code.exe"])
+
+				// Wait a moment for VS Code to close
+				await new Promise((resolve) => setTimeout(resolve, 2000))
+			} catch (taskkillError) {
+				console.warn(`Error using taskkill to quit VS Code: ${taskkillError}`)
+			}
+		} else {
+			// On Linux, try to use SIGTERM first
+			try {
+				// Find VS Code processes
+				const { stdout } = await execa("ps", ["aux"])
+				const lines = stdout.split("\n")
+
+				for (const line of lines) {
+					if (line.includes(resources.tempUserDataDir)) {
+						const parts = line.trim().split(/\s+/)
+						const pid = parseInt(parts[1])
+
+						if (pid && !isNaN(pid)) {
+							console.log(`Sending SIGTERM to VS Code process with PID: ${pid}`)
+							try {
+								// Use SIGTERM instead of SIGKILL for a graceful shutdown
+								process.kill(pid, "SIGTERM")
+							} catch (killError) {
+								console.warn(`Failed to terminate process ${pid}: ${killError}`)
+							}
+						}
+					}
+				}
+
+				// Wait a moment for VS Code to close
+				await new Promise((resolve) => setTimeout(resolve, 2000))
+			} catch (psError) {
+				console.warn(`Error listing processes: ${psError}`)
+			}
+		}
+
+		// If graceful methods failed, fall back to forceful termination as a last resort
+		// Check if VS Code is still running with the temp user data dir
+		let vsCodeStillRunning = false
+
+		if (process.platform !== "win32") {
+			try {
+				const { stdout } = await execa("ps", ["aux"])
+				vsCodeStillRunning = stdout.split("\n").some((line) => line.includes(resources.tempUserDataDir))
+			} catch (error) {
+				console.warn(`Error checking if VS Code is still running: ${error}`)
+			}
+		} else {
+			try {
+				const { stdout } = await execa("tasklist", ["/FI", `IMAGENAME eq code.exe`])
+				vsCodeStillRunning = stdout.includes("code.exe")
+			} catch (error) {
+				console.warn(`Error checking if VS Code is still running: ${error}`)
+			}
+		}
+
+		// If VS Code is still running, use forceful termination as a last resort
+		if (vsCodeStillRunning) {
+			console.log("Graceful shutdown failed, falling back to forceful termination...")
+
+			if (process.platform === "win32") {
+				try {
+					await execa("taskkill", ["/IM", "code.exe", "/F"])
+				} catch (error) {
+					console.warn(`Error forcefully terminating VS Code: ${error}`)
+				}
+			} else {
+				try {
+					const { stdout } = await execa("ps", ["aux"])
+					const lines = stdout.split("\n")
+
+					for (const line of lines) {
+						if (line.includes(resources.tempUserDataDir)) {
+							const parts = line.trim().split(/\s+/)
+							const pid = parseInt(parts[1])
+
+							if (pid && !isNaN(pid)) {
+								console.log(`Forcefully killing VS Code process with PID: ${pid}`)
+								try {
+									process.kill(pid, "SIGKILL")
+								} catch (killError) {
+									console.warn(`Failed to kill process ${pid}: ${killError}`)
+								}
+							}
+						}
+					}
+				} catch (error) {
+					console.warn(`Error forcefully terminating VS Code: ${error}`)
+				}
+			}
+		}
+	} catch (error) {
+		console.warn(`Error closing VS Code: ${error}`)
+	}
+
+	// Clean up temporary directories and evals.env file
+	try {
+		console.log(`Removing temporary user data directory: ${resources.tempUserDataDir}`)
+		fs.rmSync(resources.tempUserDataDir, { recursive: true, force: true })
+	} catch (error) {
+		console.warn(`Error removing temporary user data directory: ${error}`)
+	}
+
+	try {
+		console.log(`Removing temporary extensions directory: ${resources.tempExtensionsDir}`)
+		fs.rmSync(resources.tempExtensionsDir, { recursive: true, force: true })
+	} catch (error) {
+		console.warn(`Error removing temporary extensions directory: ${error}`)
+	}
+
+	// Remove the evals.env file
+	try {
+		const evalsEnvPath = path.join(workspacePath, "evals.env")
+		if (fs.existsSync(evalsEnvPath)) {
+			console.log(`Removing evals.env file: ${evalsEnvPath}`)
+			fs.unlinkSync(evalsEnvPath)
+		}
+	} catch (error) {
+		console.warn(`Error removing evals.env file: ${error}`)
+	}
+
+	// Remove from the global map
+	workspaceResources.delete(workspacePath)
+
+	console.log("Cleanup completed")
+}

evals/cli/tsconfig.json

@@ -0,0 +1,17 @@
+{
+	"compilerOptions": {
+		"target": "ES2020",
+		"module": "commonjs",
+		"lib": ["ES2020"],
+		"declaration": true,
+		"outDir": "dist",
+		"rootDir": "src",
+		"strict": true,
+		"esModuleInterop": true,
+		"skipLibCheck": true,
+		"forceConsistentCasingInFileNames": true,
+		"resolveJsonModule": true
+	},
+	"include": ["src/**/*"],
+	"exclude": ["node_modules", "dist"]
+}