Merge remote-tracking branch 'origin/v0.10.0rc' into stefandevo/main

apps/server/src/index.ts

@@ -55,6 +55,8 @@ import { createClaudeRoutes } from './routes/claude/index.js';
 import { ClaudeUsageService } from './services/claude-usage-service.js';
 import { createCodexRoutes } from './routes/codex/index.js';
 import { CodexUsageService } from './services/codex-usage-service.js';
+import { CodexAppServerService } from './services/codex-app-server-service.js';
+import { CodexModelCacheService } from './services/codex-model-cache-service.js';
 import { createGitHubRoutes } from './routes/github/index.js';
 import { createContextRoutes } from './routes/context/index.js';
 import { createBacklogPlanRoutes } from './routes/backlog-plan/index.js';
@@ -168,7 +170,9 @@ const agentService = new AgentService(DATA_DIR, events, settingsService);
 const featureLoader = new FeatureLoader();
 const autoModeService = new AutoModeService(events, settingsService);
 const claudeUsageService = new ClaudeUsageService();
-const codexUsageService = new CodexUsageService();
+const codexAppServerService = new CodexAppServerService();
+const codexModelCacheService = new CodexModelCacheService(DATA_DIR, codexAppServerService);
+const codexUsageService = new CodexUsageService(codexAppServerService);
 const mcpTestService = new MCPTestService(settingsService);
 const ideationService = new IdeationService(events, settingsService, featureLoader);
 
@@ -176,6 +180,11 @@ const ideationService = new IdeationService(events, settingsService, featureLoad
 (async () => {
   await agentService.initialize();
   logger.info('Agent service initialized');
+
+  // Bootstrap Codex model cache in background (don't block server startup)
+  void codexModelCacheService.getModels().catch((err) => {
+    logger.error('Failed to bootstrap Codex model cache:', err);
+  });
 })();
 
 // Run stale validation cleanup every hour to prevent memory leaks from crashed validations
@@ -219,7 +228,7 @@ app.use('/api/templates', createTemplatesRoutes());
 app.use('/api/terminal', createTerminalRoutes());
 app.use('/api/settings', createSettingsRoutes(settingsService));
 app.use('/api/claude', createClaudeRoutes(claudeUsageService));
-app.use('/api/codex', createCodexRoutes(codexUsageService));
+app.use('/api/codex', createCodexRoutes(codexUsageService, codexModelCacheService));
 app.use('/api/github', createGitHubRoutes(events, settingsService));
 app.use('/api/context', createContextRoutes(settingsService));
 app.use('/api/backlog-plan', createBacklogPlanRoutes(events, settingsService));
@@ -588,6 +597,26 @@ const startServer = (port: number) => {
 
 startServer(PORT);
 
+// Global error handlers to prevent crashes from uncaught errors
+process.on('unhandledRejection', (reason: unknown, _promise: Promise<unknown>) => {
+  logger.error('Unhandled Promise Rejection:', {
+    reason: reason instanceof Error ? reason.message : String(reason),
+    stack: reason instanceof Error ? reason.stack : undefined,
+  });
+  // Don't exit - log the error and continue running
+  // This prevents the server from crashing due to unhandled rejections
+});
+
+process.on('uncaughtException', (error: Error) => {
+  logger.error('Uncaught Exception:', {
+    message: error.message,
+    stack: error.stack,
+  });
+  // Exit on uncaught exceptions to prevent undefined behavior
+  // The process is in an unknown state after an uncaught exception
+  process.exit(1);
+});
+
 // Graceful shutdown
 process.on('SIGTERM', () => {
   logger.info('SIGTERM received, shutting down...');

apps/server/src/lib/codex-auth.ts

@@ -5,9 +5,11 @@
  * Never assumes authenticated - only returns true if CLI confirms.
  */
 
-import { spawnProcess, getCodexAuthPath } from '@automaker/platform';
+import { spawnProcess } from '@automaker/platform';
 import { findCodexCliPath } from '@automaker/platform';
-import * as fs from 'fs';
+import { createLogger } from '@automaker/utils';
+
+const logger = createLogger('CodexAuth');
 
 const CODEX_COMMAND = 'codex';
 const OPENAI_API_KEY_ENV = 'OPENAI_API_KEY';
@@ -26,36 +28,16 @@ export interface CodexAuthCheckResult {
 export async function checkCodexAuthentication(
   cliPath?: string | null
 ): Promise<CodexAuthCheckResult> {
-  console.log('[CodexAuth] checkCodexAuthentication called with cliPath:', cliPath);
-
   const resolvedCliPath = cliPath || (await findCodexCliPath());
   const hasApiKey = !!process.env[OPENAI_API_KEY_ENV];
 
-  console.log('[CodexAuth] resolvedCliPath:', resolvedCliPath);
-  console.log('[CodexAuth] hasApiKey:', hasApiKey);
-
-  // Debug: Check auth file
-  const authFilePath = getCodexAuthPath();
-  console.log('[CodexAuth] Auth file path:', authFilePath);
-  try {
-    const authFileExists = fs.existsSync(authFilePath);
-    console.log('[CodexAuth] Auth file exists:', authFileExists);
-    if (authFileExists) {
-      const authContent = fs.readFileSync(authFilePath, 'utf-8');
-      console.log('[CodexAuth] Auth file content:', authContent.substring(0, 500)); // First 500 chars
-    }
-  } catch (error) {
-    console.log('[CodexAuth] Error reading auth file:', error);
-  }
-
   // If CLI is not installed, cannot be authenticated
   if (!resolvedCliPath) {
-    console.log('[CodexAuth] No CLI path found, returning not authenticated');
+    logger.info('CLI not found');
     return { authenticated: false, method: 'none' };
   }
 
   try {
-    console.log('[CodexAuth] Running: ' + resolvedCliPath + ' login status');
     const result = await spawnProcess({
       command: resolvedCliPath || CODEX_COMMAND,
       args: ['login', 'status'],
@@ -66,33 +48,21 @@ export async function checkCodexAuthentication(
       },
     });
 
-    console.log('[CodexAuth] Command result:');
-    console.log('[CodexAuth]   exitCode:', result.exitCode);
-    console.log('[CodexAuth]   stdout:', JSON.stringify(result.stdout));
-    console.log('[CodexAuth]   stderr:', JSON.stringify(result.stderr));
-
     // Check both stdout and stderr for "logged in" - Codex CLI outputs to stderr
     const combinedOutput = (result.stdout + result.stderr).toLowerCase();
     const isLoggedIn = combinedOutput.includes('logged in');
-    console.log('[CodexAuth] isLoggedIn (contains "logged in" in stdout or stderr):', isLoggedIn);
 
     if (result.exitCode === 0 && isLoggedIn) {
       // Determine auth method based on what we know
       const method = hasApiKey ? 'api_key_env' : 'cli_authenticated';
-      console.log('[CodexAuth] Authenticated! method:', method);
+      logger.info(`✓ Authenticated (${method})`);
       return { authenticated: true, method };
     }
 
-    console.log(
-      '[CodexAuth] Not authenticated. exitCode:',
-      result.exitCode,
-      'isLoggedIn:',
-      isLoggedIn
-    );
+    logger.info('Not authenticated');
+    return { authenticated: false, method: 'none' };
   } catch (error) {
-    console.log('[CodexAuth] Error running command:', error);
+    logger.error('Failed to check authentication:', error);
+    return { authenticated: false, method: 'none' };
   }
-
-  console.log('[CodexAuth] Returning not authenticated');
-  return { authenticated: false, method: 'none' };
 }

apps/server/src/providers/codex-provider.ts

@@ -21,6 +21,7 @@ import {
   extractTextFromContent,
   classifyError,
   getUserFriendlyErrorMessage,
+  createLogger,
 } from '@automaker/utils';
 import type {
   ExecuteOptions,
@@ -658,6 +659,8 @@ async function loadCodexInstructions(cwd: string, enabled: boolean): Promise<str
     .join('\n\n');
 }
 
+const logger = createLogger('CodexProvider');
+
 export class CodexProvider extends BaseProvider {
   getName(): string {
     return 'codex';
@@ -967,21 +970,11 @@ export class CodexProvider extends BaseProvider {
   }
 
   async detectInstallation(): Promise<InstallationStatus> {
-    console.log('[CodexProvider.detectInstallation] Starting...');
-
     const cliPath = await findCodexCliPath();
     const hasApiKey = !!process.env[OPENAI_API_KEY_ENV];
     const authIndicators = await getCodexAuthIndicators();
     const installed = !!cliPath;
 
-    console.log('[CodexProvider.detectInstallation] cliPath:', cliPath);
-    console.log('[CodexProvider.detectInstallation] hasApiKey:', hasApiKey);
-    console.log(
-      '[CodexProvider.detectInstallation] authIndicators:',
-      JSON.stringify(authIndicators)
-    );
-    console.log('[CodexProvider.detectInstallation] installed:', installed);
-
     let version = '';
     if (installed) {
       try {
@@ -991,20 +984,16 @@ export class CodexProvider extends BaseProvider {
           cwd: process.cwd(),
         });
         version = result.stdout.trim();
-        console.log('[CodexProvider.detectInstallation] version:', version);
       } catch (error) {
-        console.log('[CodexProvider.detectInstallation] Error getting version:', error);
         version = '';
       }
     }
 
     // Determine auth status - always verify with CLI, never assume authenticated
-    console.log('[CodexProvider.detectInstallation] Calling checkCodexAuthentication...');
     const authCheck = await checkCodexAuthentication(cliPath);
-    console.log('[CodexProvider.detectInstallation] authCheck result:', JSON.stringify(authCheck));
     const authenticated = authCheck.authenticated;
 
-    const result = {
+    return {
       installed,
       path: cliPath || undefined,
       version: version || undefined,
@@ -1012,8 +1001,6 @@ export class CodexProvider extends BaseProvider {
       hasApiKey,
       authenticated,
     };
-    console.log('[CodexProvider.detectInstallation] Final result:', JSON.stringify(result));
-    return result;
   }
 
   getAvailableModels(): ModelDefinition[] {
@@ -1025,36 +1012,24 @@ export class CodexProvider extends BaseProvider {
    * Check authentication status for Codex CLI
    */
   async checkAuth(): Promise<CodexAuthStatus> {
-    console.log('[CodexProvider.checkAuth] Starting auth check...');
-
     const cliPath = await findCodexCliPath();
     const hasApiKey = !!process.env[OPENAI_API_KEY_ENV];
     const authIndicators = await getCodexAuthIndicators();
 
-    console.log('[CodexProvider.checkAuth] cliPath:', cliPath);
-    console.log('[CodexProvider.checkAuth] hasApiKey:', hasApiKey);
-    console.log('[CodexProvider.checkAuth] authIndicators:', JSON.stringify(authIndicators));
-
     // Check for API key in environment
     if (hasApiKey) {
-      console.log('[CodexProvider.checkAuth] Has API key, returning authenticated');
       return { authenticated: true, method: 'api_key' };
     }
 
     // Check for OAuth/token from Codex CLI
     if (authIndicators.hasOAuthToken || authIndicators.hasApiKey) {
-      console.log(
-        '[CodexProvider.checkAuth] Has OAuth token or API key in auth file, returning authenticated'
-      );
       return { authenticated: true, method: 'oauth' };
     }
 
     // CLI is installed but not authenticated via indicators - try CLI command
-    console.log('[CodexProvider.checkAuth] No indicators found, trying CLI command...');
     if (cliPath) {
       try {
         // Try 'codex login status' first (same as checkCodexAuthentication)
-        console.log('[CodexProvider.checkAuth] Running: ' + cliPath + ' login status');
         const result = await spawnProcess({
           command: cliPath || CODEX_COMMAND,
           args: ['login', 'status'],
@@ -1064,26 +1039,19 @@ export class CodexProvider extends BaseProvider {
             TERM: 'dumb',
           },
         });
-        console.log('[CodexProvider.checkAuth] login status result:');
-        console.log('[CodexProvider.checkAuth]   exitCode:', result.exitCode);
-        console.log('[CodexProvider.checkAuth]   stdout:', JSON.stringify(result.stdout));
-        console.log('[CodexProvider.checkAuth]   stderr:', JSON.stringify(result.stderr));
 
         // Check both stdout and stderr - Codex CLI outputs to stderr
         const combinedOutput = (result.stdout + result.stderr).toLowerCase();
         const isLoggedIn = combinedOutput.includes('logged in');
-        console.log('[CodexProvider.checkAuth] isLoggedIn:', isLoggedIn);
 
         if (result.exitCode === 0 && isLoggedIn) {
-          console.log('[CodexProvider.checkAuth] CLI says logged in, returning authenticated');
           return { authenticated: true, method: 'oauth' };
         }
       } catch (error) {
-        console.log('[CodexProvider.checkAuth] Error running login status:', error);
+        logger.warn('Error running login status command during auth check:', error);
       }
     }
 
-    console.log('[CodexProvider.checkAuth] Not authenticated');
     return { authenticated: false, method: 'none' };
   }
 

apps/server/src/providers/opencode-provider.ts

@@ -41,95 +41,103 @@ export interface OpenCodeAuthStatus {
 
 /**
  * Base interface for all OpenCode stream events
+ * OpenCode uses underscore format: step_start, step_finish, text
  */
 interface OpenCodeBaseEvent {
   /** Event type identifier */
   type: string;
-  /** Optional session identifier */
-  session_id?: string;
+  /** Timestamp of the event */
+  timestamp?: number;
+  /** Session ID */
+  sessionID?: string;
+  /** Part object containing the actual event data */
+  part?: Record<string, unknown>;
 }
 
 /**
- * Text delta event - Incremental text output from the model
+ * Text event - Text output from the model
+ * Format: {"type":"text","part":{"text":"content",...}}
  */
-export interface OpenCodeTextDeltaEvent extends OpenCodeBaseEvent {
-  type: 'text-delta';
-  /** The incremental text content */
-  text: string;
-}
-
-/**
- * Text end event - Signals completion of text generation
- */
-export interface OpenCodeTextEndEvent extends OpenCodeBaseEvent {
-  type: 'text-end';
+export interface OpenCodeTextEvent extends OpenCodeBaseEvent {
+  type: 'text';
+  part: {
+    type: 'text';
+    text: string;
+    [key: string]: unknown;
+  };
 }
 
 /**
  * Tool call event - Request to execute a tool
  */
 export interface OpenCodeToolCallEvent extends OpenCodeBaseEvent {
-  type: 'tool-call';
-  /** Unique identifier for this tool call */
-  call_id?: string;
-  /** Tool name to invoke */
-  name: string;
-  /** Arguments to pass to the tool */
-  args: unknown;
+  type: 'tool_call';
+  part: {
+    type: 'tool-call';
+    name: string;
+    call_id?: string;
+    args: unknown;
+    [key: string]: unknown;
+  };
 }
 
 /**
  * Tool result event - Output from a tool execution
  */
 export interface OpenCodeToolResultEvent extends OpenCodeBaseEvent {
-  type: 'tool-result';
-  /** The tool call ID this result corresponds to */
-  call_id?: string;
-  /** Output from the tool execution */
-  output: string;
+  type: 'tool_result';
+  part: {
+    type: 'tool-result';
+    call_id?: string;
+    output: string;
+    [key: string]: unknown;
+  };
 }
 
 /**
  * Tool error event - Tool execution failed
  */
 export interface OpenCodeToolErrorEvent extends OpenCodeBaseEvent {
-  type: 'tool-error';
-  /** The tool call ID that failed */
-  call_id?: string;
-  /** Error message describing the failure */
-  error: string;
+  type: 'tool_error';
+  part: {
+    type: 'tool-error';
+    call_id?: string;
+    error: string;
+    [key: string]: unknown;
+  };
 }
 
 /**
  * Start step event - Begins an agentic loop iteration
+ * Format: {"type":"step_start","part":{...}}
  */
 export interface OpenCodeStartStepEvent extends OpenCodeBaseEvent {
-  type: 'start-step';
-  /** Step number in the agentic loop */
-  step?: number;
+  type: 'step_start';
+  part?: {
+    type: 'step-start';
+    [key: string]: unknown;
+  };
 }
 
 /**
  * Finish step event - Completes an agentic loop iteration
+ * Format: {"type":"step_finish","part":{"reason":"stop",...}}
  */
 export interface OpenCodeFinishStepEvent extends OpenCodeBaseEvent {
-  type: 'finish-step';
-  /** Step number that completed */
-  step?: number;
-  /** Whether the step completed successfully */
-  success?: boolean;
-  /** Optional result data */
-  result?: string;
-  /** Optional error if step failed */
-  error?: string;
+  type: 'step_finish';
+  part?: {
+    type: 'step-finish';
+    reason?: string;
+    error?: string;
+    [key: string]: unknown;
+  };
 }
 
 /**
  * Union type of all OpenCode stream events
  */
 export type OpenCodeStreamEvent =
-  | OpenCodeTextDeltaEvent
-  | OpenCodeTextEndEvent
+  | OpenCodeTextEvent
   | OpenCodeToolCallEvent
   | OpenCodeToolResultEvent
   | OpenCodeToolErrorEvent
@@ -219,14 +227,12 @@ export class OpencodeProvider extends CliProvider {
    *
    * Arguments built:
    * - 'run' subcommand for executing queries
-   * - '--format', 'stream-json' for JSONL streaming output
-   * - '-q' / '--quiet' to suppress spinner and interactive elements
-   * - '-c', '<cwd>' for working directory
+   * - '--format', 'json' for JSON streaming output
    * - '--model', '<model>' for model selection (if specified)
-   * - '-' as final arg to read prompt from stdin
+   * - Message passed via stdin (no positional args needed)
    *
-   * The prompt is NOT included in CLI args - it's passed via stdin to avoid
-   * shell escaping issues with special characters in content.
+   * The prompt is passed via stdin to avoid shell escaping issues.
+   * OpenCode will read from stdin when no positional message arguments are provided.
    *
    * @param options - Execution options containing model, cwd, etc.
    * @returns Array of CLI arguments for opencode run
@@ -234,27 +240,18 @@ export class OpencodeProvider extends CliProvider {
   buildCliArgs(options: ExecuteOptions): string[] {
     const args: string[] = ['run'];
 
-    // Add streaming JSON output format for JSONL parsing
-    args.push('--format', 'stream-json');
-
-    // Suppress spinner and interactive elements for non-TTY usage
-    args.push('-q');
-
-    // Set working directory
-    if (options.cwd) {
-      args.push('-c', options.cwd);
-    }
+    // Add JSON output format for streaming
+    args.push('--format', 'json');
 
     // Handle model selection
-    // Strip 'opencode-' prefix if present, OpenCode uses format like 'anthropic/claude-sonnet-4-5'
+    // Strip 'opencode-' prefix if present, OpenCode uses native format
     if (options.model) {
       const model = stripProviderPrefix(options.model);
       args.push('--model', model);
     }
 
-    // Use '-' to indicate reading prompt from stdin
-    // This avoids shell escaping issues with special characters
-    args.push('-');
+    // Note: Working directory is set via subprocess cwd option, not CLI args
+    // Note: Message is passed via stdin, OpenCode reads from stdin automatically
 
     return args;
   }
@@ -314,14 +311,12 @@ export class OpencodeProvider extends CliProvider {
    * Normalize a raw CLI event to ProviderMessage format
    *
    * Maps OpenCode event types to the standard ProviderMessage structure:
-   * - text-delta -> type: 'assistant', content with type: 'text'
-   * - text-end -> null (informational, no message needed)
-   * - tool-call -> type: 'assistant', content with type: 'tool_use'
-   * - tool-result -> type: 'assistant', content with type: 'tool_result'
-   * - tool-error -> type: 'error'
-   * - start-step -> null (informational, no message needed)
-   * - finish-step with success -> type: 'result', subtype: 'success'
-   * - finish-step with error -> type: 'error'
+   * - text -> type: 'assistant', content with type: 'text'
+   * - step_start -> null (informational, no message needed)
+   * - step_finish -> type: 'result', subtype: 'success' (or error if failed)
+   * - tool_call -> type: 'assistant', content with type: 'tool_use'
+   * - tool_result -> type: 'assistant', content with type: 'tool_result'
+   * - tool_error -> type: 'error'
    *
    * @param event - Raw event from OpenCode CLI JSONL output
    * @returns Normalized ProviderMessage or null to skip the event
@@ -334,24 +329,24 @@ export class OpencodeProvider extends CliProvider {
     const openCodeEvent = event as OpenCodeStreamEvent;
 
     switch (openCodeEvent.type) {
-      case 'text-delta': {
-        const textEvent = openCodeEvent as OpenCodeTextDeltaEvent;
+      case 'text': {
+        const textEvent = openCodeEvent as OpenCodeTextEvent;
 
-        // Skip empty text deltas
-        if (!textEvent.text) {
+        // Skip if no text content
+        if (!textEvent.part?.text) {
           return null;
         }
 
         const content: ContentBlock[] = [
           {
             type: 'text',
-            text: textEvent.text,
+            text: textEvent.part.text,
           },
         ];
 
         return {
           type: 'assistant',
-          session_id: textEvent.session_id,
+          session_id: textEvent.sessionID,
           message: {
             role: 'assistant',
             content,
@@ -359,90 +354,105 @@ export class OpencodeProvider extends CliProvider {
         };
       }
 
-      case 'text-end': {
-        // Text end is informational - no message needed
-        return null;
-      }
-
-      case 'tool-call': {
-        const toolEvent = openCodeEvent as OpenCodeToolCallEvent;
-
-        // Generate a tool use ID if not provided
-        const toolUseId = toolEvent.call_id || generateToolUseId();
-
-        const content: ContentBlock[] = [
-          {
-            type: 'tool_use',
-            name: toolEvent.name,
-            tool_use_id: toolUseId,
-            input: toolEvent.args,
-          },
-        ];
-
-        return {
-          type: 'assistant',
-          session_id: toolEvent.session_id,
-          message: {
-            role: 'assistant',
-            content,
-          },
-        };
-      }
-
-      case 'tool-result': {
-        const resultEvent = openCodeEvent as OpenCodeToolResultEvent;
-
-        const content: ContentBlock[] = [
-          {
-            type: 'tool_result',
-            tool_use_id: resultEvent.call_id,
-            content: resultEvent.output,
-          },
-        ];
-
-        return {
-          type: 'assistant',
-          session_id: resultEvent.session_id,
-          message: {
-            role: 'assistant',
-            content,
-          },
-        };
-      }
-
-      case 'tool-error': {
-        const errorEvent = openCodeEvent as OpenCodeToolErrorEvent;
-
-        return {
-          type: 'error',
-          session_id: errorEvent.session_id,
-          error: errorEvent.error || 'Tool execution failed',
-        };
-      }
-
-      case 'start-step': {
+      case 'step_start': {
         // Start step is informational - no message needed
         return null;
       }
 
-      case 'finish-step': {
+      case 'step_finish': {
         const finishEvent = openCodeEvent as OpenCodeFinishStepEvent;
 
-        // Check if the step failed
-        if (finishEvent.success === false || finishEvent.error) {
+        // Check if the step failed (either has error field or reason is 'error')
+        if (finishEvent.part?.error || finishEvent.part?.reason === 'error') {
           return {
             type: 'error',
-            session_id: finishEvent.session_id,
-            error: finishEvent.error || 'Step execution failed',
+            session_id: finishEvent.sessionID,
+            error: finishEvent.part?.error || 'Step execution failed',
           };
         }
 
         // Successful completion
+        const result: { type: 'result'; subtype: 'success'; session_id?: string; result?: string } =
+          {
+            type: 'result',
+            subtype: 'success',
+          };
+
+        if (finishEvent.sessionID) {
+          result.session_id = finishEvent.sessionID;
+        }
+
+        // Safely handle arbitrary result payloads from CLI: ensure we assign a string.
+        const rawResult =
+          (finishEvent.part && (finishEvent.part as Record<string, unknown>).result) ?? undefined;
+        if (rawResult !== undefined) {
+          result.result = typeof rawResult === 'string' ? rawResult : JSON.stringify(rawResult);
+        }
+
+        return result;
+      }
+
+      case 'tool_call': {
+        const toolEvent = openCodeEvent as OpenCodeToolCallEvent;
+
+        if (!toolEvent.part) {
+          return null;
+        }
+
+        // Generate a tool use ID if not provided
+        const toolUseId = toolEvent.part.call_id || generateToolUseId();
+
+        const content: ContentBlock[] = [
+          {
+            type: 'tool_use',
+            name: toolEvent.part.name,
+            tool_use_id: toolUseId,
+            input: toolEvent.part.args,
+          },
+        ];
+
         return {
-          type: 'result',
-          subtype: 'success',
-          session_id: finishEvent.session_id,
-          result: finishEvent.result,
+          type: 'assistant',
+          session_id: toolEvent.sessionID,
+          message: {
+            role: 'assistant',
+            content,
+          },
+        };
+      }
+
+      case 'tool_result': {
+        const resultEvent = openCodeEvent as OpenCodeToolResultEvent;
+
+        if (!resultEvent.part) {
+          return null;
+        }
+
+        const content: ContentBlock[] = [
+          {
+            type: 'tool_result',
+            tool_use_id: resultEvent.part.call_id,
+            content: resultEvent.part.output,
+          },
+        ];
+
+        return {
+          type: 'assistant',
+          session_id: resultEvent.sessionID,
+          message: {
+            role: 'assistant',
+            content,
+          },
+        };
+      }
+
+      case 'tool_error': {
+        const errorEvent = openCodeEvent as OpenCodeToolErrorEvent;
+
+        return {
+          type: 'error',
+          session_id: errorEvent.sessionID,
+          error: errorEvent.part?.error || 'Tool execution failed',
         };
       }
 

apps/server/src/routes/app-spec/common.ts

@@ -6,26 +6,57 @@ import { createLogger } from '@automaker/utils';
 
 const logger = createLogger('SpecRegeneration');
 
-// Shared state for tracking generation status - private
-let isRunning = false;
-let currentAbortController: AbortController | null = null;
+// Shared state for tracking generation status - scoped by project path
+const runningProjects = new Map<string, boolean>();
+const abortControllers = new Map<string, AbortController>();
 
 /**
- * Get the current running state
+ * Get the running state for a specific project
  */
-export function getSpecRegenerationStatus(): {
+export function getSpecRegenerationStatus(projectPath?: string): {
   isRunning: boolean;
   currentAbortController: AbortController | null;
+  projectPath?: string;
 } {
-  return { isRunning, currentAbortController };
+  if (projectPath) {
+    return {
+      isRunning: runningProjects.get(projectPath) || false,
+      currentAbortController: abortControllers.get(projectPath) || null,
+      projectPath,
+    };
+  }
+  // Fallback: check if any project is running (for backward compatibility)
+  const isAnyRunning = Array.from(runningProjects.values()).some((running) => running);
+  return { isRunning: isAnyRunning, currentAbortController: null };
 }
 
 /**
- * Set the running state and abort controller
+ * Get the project path that is currently running (if any)
  */
-export function setRunningState(running: boolean, controller: AbortController | null = null): void {
-  isRunning = running;
-  currentAbortController = controller;
+export function getRunningProjectPath(): string | null {
+  for (const [path, running] of runningProjects.entries()) {
+    if (running) return path;
+  }
+  return null;
+}
+
+/**
+ * Set the running state and abort controller for a specific project
+ */
+export function setRunningState(
+  projectPath: string,
+  running: boolean,
+  controller: AbortController | null = null
+): void {
+  if (running) {
+    runningProjects.set(projectPath, true);
+    if (controller) {
+      abortControllers.set(projectPath, controller);
+    }
+  } else {
+    runningProjects.delete(projectPath);
+    abortControllers.delete(projectPath);
+  }
 }
 
 /**

apps/server/src/routes/app-spec/routes/create.ts

@@ -47,17 +47,17 @@ export function createCreateHandler(events: EventEmitter) {
         return;
       }
 
-      const { isRunning } = getSpecRegenerationStatus();
+      const { isRunning } = getSpecRegenerationStatus(projectPath);
       if (isRunning) {
-        logger.warn('Generation already running, rejecting request');
-        res.json({ success: false, error: 'Spec generation already running' });
+        logger.warn('Generation already running for project:', projectPath);
+        res.json({ success: false, error: 'Spec generation already running for this project' });
         return;
       }
 
       logAuthStatus('Before starting generation');
 
       const abortController = new AbortController();
-      setRunningState(true, abortController);
+      setRunningState(projectPath, true, abortController);
       logger.info('Starting background generation task...');
 
       // Start generation in background
@@ -80,7 +80,7 @@ export function createCreateHandler(events: EventEmitter) {
         })
         .finally(() => {
           logger.info('Generation task finished (success or error)');
-          setRunningState(false, null);
+          setRunningState(projectPath, false, null);
         });
 
       logger.info('Returning success response (generation running in background)');

apps/server/src/routes/app-spec/routes/generate-features.ts

@@ -40,17 +40,17 @@ export function createGenerateFeaturesHandler(
         return;
       }
 
-      const { isRunning } = getSpecRegenerationStatus();
+      const { isRunning } = getSpecRegenerationStatus(projectPath);
       if (isRunning) {
-        logger.warn('Generation already running, rejecting request');
-        res.json({ success: false, error: 'Generation already running' });
+        logger.warn('Generation already running for project:', projectPath);
+        res.json({ success: false, error: 'Generation already running for this project' });
         return;
       }
 
       logAuthStatus('Before starting feature generation');
 
       const abortController = new AbortController();
-      setRunningState(true, abortController);
+      setRunningState(projectPath, true, abortController);
       logger.info('Starting background feature generation task...');
 
       generateFeaturesFromSpec(projectPath, events, abortController, maxFeatures, settingsService)
@@ -63,7 +63,7 @@ export function createGenerateFeaturesHandler(
         })
         .finally(() => {
           logger.info('Feature generation task finished (success or error)');
-          setRunningState(false, null);
+          setRunningState(projectPath, false, null);
         });
 
       logger.info('Returning success response (generation running in background)');

apps/server/src/routes/app-spec/routes/generate.ts

@@ -48,17 +48,17 @@ export function createGenerateHandler(events: EventEmitter, settingsService?: Se
         return;
       }
 
-      const { isRunning } = getSpecRegenerationStatus();
+      const { isRunning } = getSpecRegenerationStatus(projectPath);
       if (isRunning) {
-        logger.warn('Generation already running, rejecting request');
-        res.json({ success: false, error: 'Spec generation already running' });
+        logger.warn('Generation already running for project:', projectPath);
+        res.json({ success: false, error: 'Spec generation already running for this project' });
         return;
       }
 
       logAuthStatus('Before starting generation');
 
       const abortController = new AbortController();
-      setRunningState(true, abortController);
+      setRunningState(projectPath, true, abortController);
       logger.info('Starting background generation task...');
 
       generateSpec(
@@ -81,7 +81,7 @@ export function createGenerateHandler(events: EventEmitter, settingsService?: Se
         })
         .finally(() => {
           logger.info('Generation task finished (success or error)');
-          setRunningState(false, null);
+          setRunningState(projectPath, false, null);
         });
 
       logger.info('Returning success response (generation running in background)');

apps/server/src/routes/app-spec/routes/status.ts

@@ -6,10 +6,11 @@ import type { Request, Response } from 'express';
 import { getSpecRegenerationStatus, getErrorMessage } from '../common.js';
 
 export function createStatusHandler() {
-  return async (_req: Request, res: Response): Promise<void> => {
+  return async (req: Request, res: Response): Promise<void> => {
     try {
-      const { isRunning } = getSpecRegenerationStatus();
-      res.json({ success: true, isRunning });
+      const projectPath = req.query.projectPath as string | undefined;
+      const { isRunning } = getSpecRegenerationStatus(projectPath);
+      res.json({ success: true, isRunning, projectPath });
     } catch (error) {
       res.status(500).json({ success: false, error: getErrorMessage(error) });
     }

apps/server/src/routes/app-spec/routes/stop.ts

@@ -6,13 +6,16 @@ import type { Request, Response } from 'express';
 import { getSpecRegenerationStatus, setRunningState, getErrorMessage } from '../common.js';
 
 export function createStopHandler() {
-  return async (_req: Request, res: Response): Promise<void> => {
+  return async (req: Request, res: Response): Promise<void> => {
     try {
-      const { currentAbortController } = getSpecRegenerationStatus();
+      const { projectPath } = req.body as { projectPath?: string };
+      const { currentAbortController } = getSpecRegenerationStatus(projectPath);
       if (currentAbortController) {
         currentAbortController.abort();
       }
-      setRunningState(false, null);
+      if (projectPath) {
+        setRunningState(projectPath, false, null);
+      }
       res.json({ success: true });
     } catch (error) {
       res.status(500).json({ success: false, error: getErrorMessage(error) });

apps/server/src/routes/auto-mode/index.ts

@@ -17,6 +17,7 @@ import { createAnalyzeProjectHandler } from './routes/analyze-project.js';
 import { createFollowUpFeatureHandler } from './routes/follow-up-feature.js';
 import { createCommitFeatureHandler } from './routes/commit-feature.js';
 import { createApprovePlanHandler } from './routes/approve-plan.js';
+import { createResumeInterruptedHandler } from './routes/resume-interrupted.js';
 
 export function createAutoModeRoutes(autoModeService: AutoModeService): Router {
   const router = Router();
@@ -63,6 +64,11 @@ export function createAutoModeRoutes(autoModeService: AutoModeService): Router {
     validatePathParams('projectPath'),
     createApprovePlanHandler(autoModeService)
   );
+  router.post(
+    '/resume-interrupted',
+    validatePathParams('projectPath'),
+    createResumeInterruptedHandler(autoModeService)
+  );
 
   return router;
 }

apps/server/src/routes/auto-mode/routes/resume-interrupted.ts

@@ -0,0 +1,42 @@
+/**
+ * Resume Interrupted Features Handler
+ *
+ * Checks for features that were interrupted (in pipeline steps or in_progress)
+ * when the server was restarted and resumes them.
+ */
+
+import type { Request, Response } from 'express';
+import { createLogger } from '@automaker/utils';
+import type { AutoModeService } from '../../../services/auto-mode-service.js';
+
+const logger = createLogger('ResumeInterrupted');
+
+interface ResumeInterruptedRequest {
+  projectPath: string;
+}
+
+export function createResumeInterruptedHandler(autoModeService: AutoModeService) {
+  return async (req: Request, res: Response): Promise<void> => {
+    const { projectPath } = req.body as ResumeInterruptedRequest;
+
+    if (!projectPath) {
+      res.status(400).json({ error: 'Project path is required' });
+      return;
+    }
+
+    logger.info(`Checking for interrupted features in ${projectPath}`);
+
+    try {
+      await autoModeService.resumeInterruptedFeatures(projectPath);
+      res.json({
+        success: true,
+        message: 'Resume check completed',
+      });
+    } catch (error) {
+      logger.error('Error resuming interrupted features:', error);
+      res.status(500).json({
+        error: error instanceof Error ? error.message : 'Unknown error',
+      });
+    }
+  };
+}

apps/server/src/routes/codex/index.ts

@@ -1,17 +1,21 @@
 import { Router, Request, Response } from 'express';
 import { CodexUsageService } from '../../services/codex-usage-service.js';
+import { CodexModelCacheService } from '../../services/codex-model-cache-service.js';
 import { createLogger } from '@automaker/utils';
 
 const logger = createLogger('Codex');
 
-export function createCodexRoutes(service: CodexUsageService): Router {
+export function createCodexRoutes(
+  usageService: CodexUsageService,
+  modelCacheService: CodexModelCacheService
+): Router {
   const router = Router();
 
   // Get current usage (attempts to fetch from Codex CLI)
-  router.get('/usage', async (req: Request, res: Response) => {
+  router.get('/usage', async (_req: Request, res: Response) => {
     try {
       // Check if Codex CLI is available first
-      const isAvailable = await service.isAvailable();
+      const isAvailable = await usageService.isAvailable();
       if (!isAvailable) {
         // IMPORTANT: This endpoint is behind Automaker session auth already.
         // Use a 200 + error payload for Codex CLI issues so the UI doesn't
@@ -23,7 +27,7 @@ export function createCodexRoutes(service: CodexUsageService): Router {
         return;
       }
 
-      const usage = await service.fetchUsageData();
+      const usage = await usageService.fetchUsageData();
       res.json(usage);
     } catch (error) {
       const message = error instanceof Error ? error.message : 'Unknown error';
@@ -52,5 +56,35 @@ export function createCodexRoutes(service: CodexUsageService): Router {
     }
   });
 
+  // Get available Codex models (cached)
+  router.get('/models', async (req: Request, res: Response) => {
+    try {
+      const forceRefresh = req.query.refresh === 'true';
+      const { models, cachedAt } = await modelCacheService.getModelsWithMetadata(forceRefresh);
+
+      if (models.length === 0) {
+        res.status(503).json({
+          success: false,
+          error: 'Codex CLI not available or not authenticated',
+          message: "Please install Codex CLI and run 'codex login' to authenticate",
+        });
+        return;
+      }
+
+      res.json({
+        success: true,
+        models,
+        cachedAt,
+      });
+    } catch (error) {
+      logger.error('Error fetching models:', error);
+      const message = error instanceof Error ? error.message : 'Unknown error';
+      res.status(500).json({
+        success: false,
+        error: message,
+      });
+    }
+  });
+
   return router;
 }

apps/server/src/routes/enhance-prompt/routes/enhance.ts

@@ -188,6 +188,7 @@ export function createEnhanceHandler(
         technical: prompts.enhancement.technicalSystemPrompt,
         simplify: prompts.enhancement.simplifySystemPrompt,
         acceptance: prompts.enhancement.acceptanceSystemPrompt,
+        'ux-reviewer': prompts.enhancement.uxReviewerSystemPrompt,
       };
       const systemPrompt = systemPromptMap[validMode];
 

apps/server/src/routes/features/index.ts

@@ -10,6 +10,7 @@ import { createGetHandler } from './routes/get.js';
 import { createCreateHandler } from './routes/create.js';
 import { createUpdateHandler } from './routes/update.js';
 import { createBulkUpdateHandler } from './routes/bulk-update.js';
+import { createBulkDeleteHandler } from './routes/bulk-delete.js';
 import { createDeleteHandler } from './routes/delete.js';
 import { createAgentOutputHandler, createRawOutputHandler } from './routes/agent-output.js';
 import { createGenerateTitleHandler } from './routes/generate-title.js';
@@ -26,6 +27,11 @@ export function createFeaturesRoutes(featureLoader: FeatureLoader): Router {
     validatePathParams('projectPath'),
     createBulkUpdateHandler(featureLoader)
   );
+  router.post(
+    '/bulk-delete',
+    validatePathParams('projectPath'),
+    createBulkDeleteHandler(featureLoader)
+  );
   router.post('/delete', validatePathParams('projectPath'), createDeleteHandler(featureLoader));
   router.post('/agent-output', createAgentOutputHandler(featureLoader));
   router.post('/raw-output', createRawOutputHandler(featureLoader));

apps/server/src/routes/features/routes/bulk-delete.ts

@@ -0,0 +1,61 @@
+/**
+ * POST /bulk-delete endpoint - Delete multiple features at once
+ */
+
+import type { Request, Response } from 'express';
+import { FeatureLoader } from '../../../services/feature-loader.js';
+import { getErrorMessage, logError } from '../common.js';
+
+interface BulkDeleteRequest {
+  projectPath: string;
+  featureIds: string[];
+}
+
+interface BulkDeleteResult {
+  featureId: string;
+  success: boolean;
+  error?: string;
+}
+
+export function createBulkDeleteHandler(featureLoader: FeatureLoader) {
+  return async (req: Request, res: Response): Promise<void> => {
+    try {
+      const { projectPath, featureIds } = req.body as BulkDeleteRequest;
+
+      if (!projectPath || !featureIds || !Array.isArray(featureIds) || featureIds.length === 0) {
+        res.status(400).json({
+          success: false,
+          error: 'projectPath and featureIds (non-empty array) are required',
+        });
+        return;
+      }
+
+      const results = await Promise.all(
+        featureIds.map(async (featureId) => {
+          const success = await featureLoader.delete(projectPath, featureId);
+          if (success) {
+            return { featureId, success: true };
+          }
+          return {
+            featureId,
+            success: false,
+            error: 'Deletion failed. Check server logs for details.',
+          };
+        })
+      );
+
+      const successCount = results.reduce((count, r) => count + (r.success ? 1 : 0), 0);
+      const failureCount = results.length - successCount;
+
+      res.json({
+        success: failureCount === 0,
+        deletedCount: successCount,
+        failedCount: failureCount,
+        results,
+      });
+    } catch (error) {
+      logError(error, 'Bulk delete features failed');
+      res.status(500).json({ success: false, error: getErrorMessage(error) });
+    }
+  };
+}

apps/server/src/routes/features/routes/update.ts

@@ -10,14 +10,21 @@ import { getErrorMessage, logError } from '../common.js';
 export function createUpdateHandler(featureLoader: FeatureLoader) {
   return async (req: Request, res: Response): Promise<void> => {
     try {
-      const { projectPath, featureId, updates, descriptionHistorySource, enhancementMode } =
-        req.body as {
-          projectPath: string;
-          featureId: string;
-          updates: Partial<Feature>;
-          descriptionHistorySource?: 'enhance' | 'edit';
-          enhancementMode?: 'improve' | 'technical' | 'simplify' | 'acceptance';
-        };
+      const {
+        projectPath,
+        featureId,
+        updates,
+        descriptionHistorySource,
+        enhancementMode,
+        preEnhancementDescription,
+      } = req.body as {
+        projectPath: string;
+        featureId: string;
+        updates: Partial<Feature>;
+        descriptionHistorySource?: 'enhance' | 'edit';
+        enhancementMode?: 'improve' | 'technical' | 'simplify' | 'acceptance' | 'ux-reviewer';
+        preEnhancementDescription?: string;
+      };
 
       if (!projectPath || !featureId || !updates) {
         res.status(400).json({
@@ -32,7 +39,8 @@ export function createUpdateHandler(featureLoader: FeatureLoader) {
         featureId,
         updates,
         descriptionHistorySource,
-        enhancementMode
+        enhancementMode,
+        preEnhancementDescription
       );
       res.json({ success: true, feature: updated });
     } catch (error) {

apps/server/src/routes/worktree/routes/list.ts

@@ -2,18 +2,23 @@
  * POST /list endpoint - List all git worktrees
  *
  * Returns actual git worktrees from `git worktree list`.
+ * Also scans .worktrees/ directory to discover worktrees that may have been
+ * created externally or whose git state was corrupted.
  * Does NOT include tracked branches - only real worktrees with separate directories.
  */
 
 import type { Request, Response } from 'express';
 import { exec } from 'child_process';
 import { promisify } from 'util';
+import path from 'path';
 import * as secureFs from '../../../lib/secure-fs.js';
 import { isGitRepo } from '@automaker/git-utils';
 import { getErrorMessage, logError, normalizePath } from '../common.js';
 import { readAllWorktreeMetadata, type WorktreePRInfo } from '../../../lib/worktree-metadata.js';
+import { createLogger } from '@automaker/utils';
 
 const execAsync = promisify(exec);
+const logger = createLogger('Worktree');
 
 interface WorktreeInfo {
   path: string;
@@ -35,6 +40,87 @@ async function getCurrentBranch(cwd: string): Promise<string> {
   }
 }
 
+/**
+ * Scan the .worktrees directory to discover worktrees that may exist on disk
+ * but are not registered with git (e.g., created externally or corrupted state).
+ */
+async function scanWorktreesDirectory(
+  projectPath: string,
+  knownWorktreePaths: Set<string>
+): Promise<Array<{ path: string; branch: string }>> {
+  const discovered: Array<{ path: string; branch: string }> = [];
+  const worktreesDir = path.join(projectPath, '.worktrees');
+
+  try {
+    // Check if .worktrees directory exists
+    await secureFs.access(worktreesDir);
+  } catch {
+    // .worktrees directory doesn't exist
+    return discovered;
+  }
+
+  try {
+    const entries = await secureFs.readdir(worktreesDir, { withFileTypes: true });
+
+    for (const entry of entries) {
+      if (!entry.isDirectory()) continue;
+
+      const worktreePath = path.join(worktreesDir, entry.name);
+      const normalizedPath = normalizePath(worktreePath);
+
+      // Skip if already known from git worktree list
+      if (knownWorktreePaths.has(normalizedPath)) continue;
+
+      // Check if this is a valid git repository
+      const gitPath = path.join(worktreePath, '.git');
+      try {
+        const gitStat = await secureFs.stat(gitPath);
+
+        // Git worktrees have a .git FILE (not directory) that points to the parent repo
+        // Regular repos have a .git DIRECTORY
+        if (gitStat.isFile() || gitStat.isDirectory()) {
+          // Try to get the branch name
+          const branch = await getCurrentBranch(worktreePath);
+          if (branch) {
+            logger.info(
+              `Discovered worktree in .worktrees/ not in git worktree list: ${entry.name} (branch: ${branch})`
+            );
+            discovered.push({
+              path: normalizedPath,
+              branch,
+            });
+          } else {
+            // Try to get branch from HEAD if branch --show-current fails (detached HEAD)
+            try {
+              const { stdout: headRef } = await execAsync('git rev-parse --abbrev-ref HEAD', {
+                cwd: worktreePath,
+              });
+              const headBranch = headRef.trim();
+              if (headBranch && headBranch !== 'HEAD') {
+                logger.info(
+                  `Discovered worktree in .worktrees/ not in git worktree list: ${entry.name} (branch: ${headBranch})`
+                );
+                discovered.push({
+                  path: normalizedPath,
+                  branch: headBranch,
+                });
+              }
+            } catch {
+              // Can't determine branch, skip this directory
+            }
+          }
+        }
+      } catch {
+        // Not a git repo, skip
+      }
+    }
+  } catch (error) {
+    logger.warn(`Failed to scan .worktrees directory: ${getErrorMessage(error)}`);
+  }
+
+  return discovered;
+}
+
 export function createListHandler() {
   return async (req: Request, res: Response): Promise<void> => {
     try {
@@ -116,6 +202,22 @@ export function createListHandler() {
         }
       }
 
+      // Scan .worktrees directory to discover worktrees that exist on disk
+      // but are not registered with git (e.g., created externally)
+      const knownPaths = new Set(worktrees.map((w) => w.path));
+      const discoveredWorktrees = await scanWorktreesDirectory(projectPath, knownPaths);
+
+      // Add discovered worktrees to the list
+      for (const discovered of discoveredWorktrees) {
+        worktrees.push({
+          path: discovered.path,
+          branch: discovered.branch,
+          isMain: false,
+          isCurrent: discovered.branch === currentBranch,
+          hasWorktree: true,
+        });
+      }
+
       // Read all worktree metadata to get PR info
       const allMetadata = await readAllWorktreeMetadata(projectPath);
 

apps/server/src/services/auto-mode-service.ts

@@ -31,7 +31,13 @@ import {
 const logger = createLogger('AutoMode');
 import { resolveModelString, resolvePhaseModel, DEFAULT_MODELS } from '@automaker/model-resolver';
 import { resolveDependencies, areDependenciesSatisfied } from '@automaker/dependency-resolver';
-import { getFeatureDir, getAutomakerDir, getFeaturesDir } from '@automaker/platform';
+import {
+  getFeatureDir,
+  getAutomakerDir,
+  getFeaturesDir,
+  getExecutionStatePath,
+  ensureAutomakerDir,
+} from '@automaker/platform';
 import { exec } from 'child_process';
 import { promisify } from 'util';
 import path from 'path';
@@ -201,6 +207,29 @@ interface AutoModeConfig {
   projectPath: string;
 }
 
+/**
+ * Execution state for recovery after server restart
+ * Tracks which features were running and auto-loop configuration
+ */
+interface ExecutionState {
+  version: 1;
+  autoLoopWasRunning: boolean;
+  maxConcurrency: number;
+  projectPath: string;
+  runningFeatureIds: string[];
+  savedAt: string;
+}
+
+// Default empty execution state
+const DEFAULT_EXECUTION_STATE: ExecutionState = {
+  version: 1,
+  autoLoopWasRunning: false,
+  maxConcurrency: 3,
+  projectPath: '',
+  runningFeatureIds: [],
+  savedAt: '',
+};
+
 // Constants for consecutive failure tracking
 const CONSECUTIVE_FAILURE_THRESHOLD = 3; // Pause after 3 consecutive failures
 const FAILURE_WINDOW_MS = 60000; // Failures within 1 minute count as consecutive
@@ -322,6 +351,9 @@ export class AutoModeService {
       projectPath,
     });
 
+    // Save execution state for recovery after restart
+    await this.saveExecutionState(projectPath);
+
     // Note: Memory folder initialization is now handled by loadContextFiles
 
     // Run the loop in the background
@@ -390,17 +422,23 @@ export class AutoModeService {
    */
   async stopAutoLoop(): Promise<number> {
     const wasRunning = this.autoLoopRunning;
+    const projectPath = this.config?.projectPath;
     this.autoLoopRunning = false;
     if (this.autoLoopAbortController) {
       this.autoLoopAbortController.abort();
       this.autoLoopAbortController = null;
     }
 
+    // Clear execution state when auto-loop is explicitly stopped
+    if (projectPath) {
+      await this.clearExecutionState(projectPath);
+    }
+
     // Emit stop event immediately when user explicitly stops
     if (wasRunning) {
       this.emitAutoModeEvent('auto_mode_stopped', {
         message: 'Auto mode stopped',
-        projectPath: this.config?.projectPath,
+        projectPath,
       });
     }
 
@@ -441,6 +479,11 @@ export class AutoModeService {
     };
     this.runningFeatures.set(featureId, tempRunningFeature);
 
+    // Save execution state when feature starts
+    if (isAutoMode) {
+      await this.saveExecutionState(projectPath);
+    }
+
     try {
       // Validate that project path is allowed using centralized validation
       validateWorkingDirectory(projectPath);
@@ -695,6 +738,11 @@ export class AutoModeService {
         `Pending approvals at cleanup: ${Array.from(this.pendingApprovals.keys()).join(', ') || 'none'}`
       );
       this.runningFeatures.delete(featureId);
+
+      // Update execution state after feature completes
+      if (this.autoLoopRunning && projectPath) {
+        await this.saveExecutionState(projectPath);
+      }
     }
   }
 
@@ -2950,6 +2998,149 @@ Begin implementing task ${task.id} now.`;
     });
   }
 
+  // ============================================================================
+  // Execution State Persistence - For recovery after server restart
+  // ============================================================================
+
+  /**
+   * Save execution state to disk for recovery after server restart
+   */
+  private async saveExecutionState(projectPath: string): Promise<void> {
+    try {
+      await ensureAutomakerDir(projectPath);
+      const statePath = getExecutionStatePath(projectPath);
+      const state: ExecutionState = {
+        version: 1,
+        autoLoopWasRunning: this.autoLoopRunning,
+        maxConcurrency: this.config?.maxConcurrency ?? 3,
+        projectPath,
+        runningFeatureIds: Array.from(this.runningFeatures.keys()),
+        savedAt: new Date().toISOString(),
+      };
+      await secureFs.writeFile(statePath, JSON.stringify(state, null, 2), 'utf-8');
+      logger.info(`Saved execution state: ${state.runningFeatureIds.length} running features`);
+    } catch (error) {
+      logger.error('Failed to save execution state:', error);
+    }
+  }
+
+  /**
+   * Load execution state from disk
+   */
+  private async loadExecutionState(projectPath: string): Promise<ExecutionState> {
+    try {
+      const statePath = getExecutionStatePath(projectPath);
+      const content = (await secureFs.readFile(statePath, 'utf-8')) as string;
+      const state = JSON.parse(content) as ExecutionState;
+      return state;
+    } catch (error) {
+      if ((error as NodeJS.ErrnoException).code !== 'ENOENT') {
+        logger.error('Failed to load execution state:', error);
+      }
+      return DEFAULT_EXECUTION_STATE;
+    }
+  }
+
+  /**
+   * Clear execution state (called on successful shutdown or when auto-loop stops)
+   */
+  private async clearExecutionState(projectPath: string): Promise<void> {
+    try {
+      const statePath = getExecutionStatePath(projectPath);
+      await secureFs.unlink(statePath);
+      logger.info('Cleared execution state');
+    } catch (error) {
+      if ((error as NodeJS.ErrnoException).code !== 'ENOENT') {
+        logger.error('Failed to clear execution state:', error);
+      }
+    }
+  }
+
+  /**
+   * Check for and resume interrupted features after server restart
+   * This should be called during server initialization
+   */
+  async resumeInterruptedFeatures(projectPath: string): Promise<void> {
+    logger.info('Checking for interrupted features to resume...');
+
+    // Load all features and find those that were interrupted
+    const featuresDir = getFeaturesDir(projectPath);
+
+    try {
+      const entries = await secureFs.readdir(featuresDir, { withFileTypes: true });
+      const interruptedFeatures: Feature[] = [];
+
+      for (const entry of entries) {
+        if (entry.isDirectory()) {
+          const featurePath = path.join(featuresDir, entry.name, 'feature.json');
+          try {
+            const data = (await secureFs.readFile(featurePath, 'utf-8')) as string;
+            const feature = JSON.parse(data) as Feature;
+
+            // Check if feature was interrupted (in_progress or pipeline_*)
+            if (
+              feature.status === 'in_progress' ||
+              (feature.status && feature.status.startsWith('pipeline_'))
+            ) {
+              // Verify it has existing context (agent-output.md)
+              const featureDir = getFeatureDir(projectPath, feature.id);
+              const contextPath = path.join(featureDir, 'agent-output.md');
+              try {
+                await secureFs.access(contextPath);
+                interruptedFeatures.push(feature);
+                logger.info(
+                  `Found interrupted feature: ${feature.id} (${feature.title}) - status: ${feature.status}`
+                );
+              } catch {
+                // No context file, skip this feature - it will be restarted fresh
+                logger.info(`Interrupted feature ${feature.id} has no context, will restart fresh`);
+              }
+            }
+          } catch {
+            // Skip invalid features
+          }
+        }
+      }
+
+      if (interruptedFeatures.length === 0) {
+        logger.info('No interrupted features found');
+        return;
+      }
+
+      logger.info(`Found ${interruptedFeatures.length} interrupted feature(s) to resume`);
+
+      // Emit event to notify UI
+      this.emitAutoModeEvent('auto_mode_resuming_features', {
+        message: `Resuming ${interruptedFeatures.length} interrupted feature(s) after server restart`,
+        projectPath,
+        featureIds: interruptedFeatures.map((f) => f.id),
+        features: interruptedFeatures.map((f) => ({
+          id: f.id,
+          title: f.title,
+          status: f.status,
+        })),
+      });
+
+      // Resume each interrupted feature
+      for (const feature of interruptedFeatures) {
+        try {
+          logger.info(`Resuming feature: ${feature.id} (${feature.title})`);
+          // Use resumeFeature which will detect the existing context and continue
+          await this.resumeFeature(projectPath, feature.id, true);
+        } catch (error) {
+          logger.error(`Failed to resume feature ${feature.id}:`, error);
+          // Continue with other features
+        }
+      }
+    } catch (error) {
+      if ((error as NodeJS.ErrnoException).code === 'ENOENT') {
+        logger.info('No features directory found, nothing to resume');
+      } else {
+        logger.error('Error checking for interrupted features:', error);
+      }
+    }
+  }
+
   /**
    * Extract and record learnings from a completed feature
    * Uses a quick Claude call to identify important decisions and patterns

apps/server/src/services/claude-usage-service.ts

@@ -2,6 +2,7 @@ import { spawn } from 'child_process';
 import * as os from 'os';
 import * as pty from 'node-pty';
 import { ClaudeUsage } from '../routes/claude/types.js';
+import { createLogger } from '@automaker/utils';
 
 /**
  * Claude Usage Service
@@ -14,6 +15,8 @@ import { ClaudeUsage } from '../routes/claude/types.js';
  * - macOS: Uses 'expect' command for PTY
  * - Windows/Linux: Uses node-pty for PTY
  */
+const logger = createLogger('ClaudeUsage');
+
 export class ClaudeUsageService {
   private claudeBinary = 'claude';
   private timeout = 30000; // 30 second timeout
@@ -164,21 +167,40 @@ export class ClaudeUsageService {
       const shell = this.isWindows ? 'cmd.exe' : '/bin/sh';
       const args = this.isWindows ? ['/c', 'claude', '/usage'] : ['-c', 'claude /usage'];
 
-      const ptyProcess = pty.spawn(shell, args, {
-        name: 'xterm-256color',
-        cols: 120,
-        rows: 30,
-        cwd: workingDirectory,
-        env: {
-          ...process.env,
-          TERM: 'xterm-256color',
-        } as Record<string, string>,
-      });
+      let ptyProcess: any = null;
+
+      try {
+        ptyProcess = pty.spawn(shell, args, {
+          name: 'xterm-256color',
+          cols: 120,
+          rows: 30,
+          cwd: workingDirectory,
+          env: {
+            ...process.env,
+            TERM: 'xterm-256color',
+          } as Record<string, string>,
+        });
+      } catch (spawnError) {
+        // pty.spawn() can throw synchronously if the native module fails to load
+        // or if PTY is not available in the current environment (e.g., containers without /dev/pts)
+        const errorMessage = spawnError instanceof Error ? spawnError.message : String(spawnError);
+        logger.error('[executeClaudeUsageCommandPty] Failed to spawn PTY:', errorMessage);
+
+        // Return a user-friendly error instead of crashing
+        reject(
+          new Error(
+            `Unable to access terminal: ${errorMessage}. Claude CLI may not be available or PTY support is limited in this environment.`
+          )
+        );
+        return;
+      }
 
       const timeoutId = setTimeout(() => {
         if (!settled) {
           settled = true;
-          ptyProcess.kill();
+          if (ptyProcess && !ptyProcess.killed) {
+            ptyProcess.kill();
+          }
           // Don't fail if we have data - return it instead
           if (output.includes('Current session')) {
             resolve(output);
@@ -188,7 +210,7 @@ export class ClaudeUsageService {
         }
       }, this.timeout);
 
-      ptyProcess.onData((data) => {
+      ptyProcess.onData((data: string) => {
         output += data;
 
         // Check if we've seen the usage data (look for "Current session")
@@ -196,12 +218,12 @@ export class ClaudeUsageService {
           hasSeenUsageData = true;
           // Wait for full output, then send escape to exit
           setTimeout(() => {
-            if (!settled) {
+            if (!settled && ptyProcess && !ptyProcess.killed) {
               ptyProcess.write('\x1b'); // Send escape key
 
               // Fallback: if ESC doesn't exit (Linux), use SIGTERM after 2s
               setTimeout(() => {
-                if (!settled) {
+                if (!settled && ptyProcess && !ptyProcess.killed) {
                   ptyProcess.kill('SIGTERM');
                 }
               }, 2000);
@@ -212,14 +234,14 @@ export class ClaudeUsageService {
         // Fallback: if we see "Esc to cancel" but haven't seen usage data yet
         if (!hasSeenUsageData && output.includes('Esc to cancel')) {
           setTimeout(() => {
-            if (!settled) {
+            if (!settled && ptyProcess && !ptyProcess.killed) {
               ptyProcess.write('\x1b'); // Send escape key
             }
           }, 3000);
         }
       });
 
-      ptyProcess.onExit(({ exitCode }) => {
+      ptyProcess.onExit(({ exitCode }: { exitCode: number }) => {
         clearTimeout(timeoutId);
         if (settled) return;
         settled = true;

apps/server/src/services/codex-app-server-service.ts

@@ -0,0 +1,212 @@
+import { spawn, type ChildProcess } from 'child_process';
+import readline from 'readline';
+import { findCodexCliPath } from '@automaker/platform';
+import { createLogger } from '@automaker/utils';
+import type {
+  AppServerModelResponse,
+  AppServerAccountResponse,
+  AppServerRateLimitsResponse,
+  JsonRpcRequest,
+} from '@automaker/types';
+
+const logger = createLogger('CodexAppServer');
+
+/**
+ * CodexAppServerService
+ *
+ * Centralized service for communicating with Codex CLI's app-server via JSON-RPC protocol.
+ * Handles process spawning, JSON-RPC messaging, and cleanup.
+ *
+ * Connection strategy: Spawn on-demand (new process for each method call)
+ */
+export class CodexAppServerService {
+  private cachedCliPath: string | null = null;
+
+  /**
+   * Check if Codex CLI is available on the system
+   */
+  async isAvailable(): Promise<boolean> {
+    this.cachedCliPath = await findCodexCliPath();
+    return Boolean(this.cachedCliPath);
+  }
+
+  /**
+   * Fetch available models from app-server
+   */
+  async getModels(): Promise<AppServerModelResponse | null> {
+    const result = await this.executeJsonRpc<AppServerModelResponse>((sendRequest) => {
+      return sendRequest('model/list', {});
+    });
+
+    if (result) {
+      logger.info(`[getModels] ✓ Fetched ${result.data.length} models`);
+    }
+
+    return result;
+  }
+
+  /**
+   * Fetch account information from app-server
+   */
+  async getAccount(): Promise<AppServerAccountResponse | null> {
+    return this.executeJsonRpc<AppServerAccountResponse>((sendRequest) => {
+      return sendRequest('account/read', { refreshToken: false });
+    });
+  }
+
+  /**
+   * Fetch rate limits from app-server
+   */
+  async getRateLimits(): Promise<AppServerRateLimitsResponse | null> {
+    return this.executeJsonRpc<AppServerRateLimitsResponse>((sendRequest) => {
+      return sendRequest('account/rateLimits/read', {});
+    });
+  }
+
+  /**
+   * Execute JSON-RPC requests via Codex app-server
+   *
+   * This method:
+   * 1. Spawns a new `codex app-server` process
+   * 2. Handles JSON-RPC initialization handshake
+   * 3. Executes user-provided requests
+   * 4. Cleans up the process
+   *
+   * @param requestFn - Function that receives sendRequest helper and returns a promise
+   * @returns Result of the JSON-RPC request or null on failure
+   */
+  private async executeJsonRpc<T>(
+    requestFn: (sendRequest: <R>(method: string, params?: unknown) => Promise<R>) => Promise<T>
+  ): Promise<T | null> {
+    let childProcess: ChildProcess | null = null;
+
+    try {
+      const cliPath = this.cachedCliPath || (await findCodexCliPath());
+
+      if (!cliPath) {
+        return null;
+      }
+
+      // On Windows, .cmd files must be run through shell
+      const needsShell = process.platform === 'win32' && cliPath.toLowerCase().endsWith('.cmd');
+
+      childProcess = spawn(cliPath, ['app-server'], {
+        cwd: process.cwd(),
+        env: {
+          ...process.env,
+          TERM: 'dumb',
+        },
+        stdio: ['pipe', 'pipe', 'pipe'],
+        shell: needsShell,
+      });
+
+      if (!childProcess.stdin || !childProcess.stdout) {
+        throw new Error('Failed to create stdio pipes');
+      }
+
+      // Setup readline for reading JSONL responses
+      const rl = readline.createInterface({
+        input: childProcess.stdout,
+        crlfDelay: Infinity,
+      });
+
+      // Message ID counter for JSON-RPC
+      let messageId = 0;
+      const pendingRequests = new Map<
+        number,
+        {
+          resolve: (value: unknown) => void;
+          reject: (error: Error) => void;
+          timeout: NodeJS.Timeout;
+        }
+      >();
+
+      // Process incoming messages
+      rl.on('line', (line) => {
+        if (!line.trim()) return;
+
+        try {
+          const message = JSON.parse(line);
+
+          // Handle response to our request
+          if ('id' in message && message.id !== undefined) {
+            const pending = pendingRequests.get(message.id);
+            if (pending) {
+              clearTimeout(pending.timeout);
+              pendingRequests.delete(message.id);
+              if (message.error) {
+                pending.reject(new Error(message.error.message || 'Unknown error'));
+              } else {
+                pending.resolve(message.result);
+              }
+            }
+          }
+          // Ignore notifications (no id field)
+        } catch {
+          // Ignore parse errors for non-JSON lines
+        }
+      });
+
+      // Helper to send JSON-RPC request and wait for response
+      const sendRequest = <R>(method: string, params?: unknown): Promise<R> => {
+        return new Promise((resolve, reject) => {
+          const id = ++messageId;
+          const request: JsonRpcRequest = {
+            method,
+            id,
+            params: params ?? {},
+          };
+
+          // Set timeout for request (10 seconds)
+          const timeout = setTimeout(() => {
+            pendingRequests.delete(id);
+            reject(new Error(`Request timeout: ${method}`));
+          }, 10000);
+
+          pendingRequests.set(id, {
+            resolve: resolve as (value: unknown) => void,
+            reject,
+            timeout,
+          });
+
+          childProcess!.stdin!.write(JSON.stringify(request) + '\n');
+        });
+      };
+
+      // Helper to send notification (no response expected)
+      const sendNotification = (method: string, params?: unknown): void => {
+        const notification = params ? { method, params } : { method };
+        childProcess!.stdin!.write(JSON.stringify(notification) + '\n');
+      };
+
+      // 1. Initialize the app-server
+      await sendRequest('initialize', {
+        clientInfo: {
+          name: 'automaker',
+          title: 'AutoMaker',
+          version: '1.0.0',
+        },
+      });
+
+      // 2. Send initialized notification
+      sendNotification('initialized');
+
+      // 3. Execute user-provided requests
+      const result = await requestFn(sendRequest);
+
+      // Clean up
+      rl.close();
+      childProcess.kill('SIGTERM');
+
+      return result;
+    } catch (error) {
+      logger.error('[executeJsonRpc] Failed:', error);
+      return null;
+    } finally {
+      // Ensure process is killed
+      if (childProcess && !childProcess.killed) {
+        childProcess.kill('SIGTERM');
+      }
+    }
+  }
+}

apps/server/src/services/codex-model-cache-service.ts

@@ -0,0 +1,258 @@
+import path from 'path';
+import { secureFs } from '@automaker/platform';
+import { createLogger } from '@automaker/utils';
+import type { AppServerModel } from '@automaker/types';
+import type { CodexAppServerService } from './codex-app-server-service.js';
+
+const logger = createLogger('CodexModelCache');
+
+/**
+ * Codex model with UI-compatible format
+ */
+export interface CodexModel {
+  id: string;
+  label: string;
+  description: string;
+  hasThinking: boolean;
+  supportsVision: boolean;
+  tier: 'premium' | 'standard' | 'basic';
+  isDefault: boolean;
+}
+
+/**
+ * Cache structure stored on disk
+ */
+interface CodexModelCache {
+  models: CodexModel[];
+  cachedAt: number;
+  ttl: number;
+}
+
+/**
+ * CodexModelCacheService
+ *
+ * Caches Codex models fetched from app-server with TTL-based invalidation and disk persistence.
+ *
+ * Features:
+ * - 1-hour TTL (configurable)
+ * - Atomic file writes (temp file + rename)
+ * - Thread-safe (deduplicates concurrent refresh requests)
+ * - Auto-bootstrap on service creation
+ * - Graceful fallback (returns empty array on errors)
+ */
+export class CodexModelCacheService {
+  private cacheFilePath: string;
+  private ttl: number;
+  private appServerService: CodexAppServerService;
+  private inFlightRefresh: Promise<CodexModel[]> | null = null;
+
+  constructor(
+    dataDir: string,
+    appServerService: CodexAppServerService,
+    ttl: number = 3600000 // 1 hour default
+  ) {
+    this.cacheFilePath = path.join(dataDir, 'codex-models-cache.json');
+    this.ttl = ttl;
+    this.appServerService = appServerService;
+  }
+
+  /**
+   * Get models from cache or fetch if stale
+   *
+   * @param forceRefresh - If true, bypass cache and fetch fresh data
+   * @returns Array of Codex models (empty array if unavailable)
+   */
+  async getModels(forceRefresh = false): Promise<CodexModel[]> {
+    // If force refresh, skip cache
+    if (forceRefresh) {
+      return this.refreshModels();
+    }
+
+    // Try to load from cache
+    const cached = await this.loadFromCache();
+    if (cached) {
+      const age = Date.now() - cached.cachedAt;
+      const isStale = age > cached.ttl;
+
+      if (!isStale) {
+        logger.info(
+          `[getModels] ✓ Using cached models (${cached.models.length} models, age: ${Math.round(age / 60000)}min)`
+        );
+        return cached.models;
+      }
+    }
+
+    // Cache is stale or missing, refresh
+    return this.refreshModels();
+  }
+
+  /**
+   * Get models with cache metadata
+   *
+   * @param forceRefresh - If true, bypass cache and fetch fresh data
+   * @returns Object containing models and cache timestamp
+   */
+  async getModelsWithMetadata(
+    forceRefresh = false
+  ): Promise<{ models: CodexModel[]; cachedAt: number }> {
+    const models = await this.getModels(forceRefresh);
+
+    // Try to get the actual cache timestamp
+    const cached = await this.loadFromCache();
+    const cachedAt = cached?.cachedAt ?? Date.now();
+
+    return { models, cachedAt };
+  }
+
+  /**
+   * Refresh models from app-server and update cache
+   *
+   * Thread-safe: Deduplicates concurrent refresh requests
+   */
+  async refreshModels(): Promise<CodexModel[]> {
+    // Deduplicate concurrent refresh requests
+    if (this.inFlightRefresh) {
+      return this.inFlightRefresh;
+    }
+
+    // Start new refresh
+    this.inFlightRefresh = this.doRefresh();
+
+    try {
+      const models = await this.inFlightRefresh;
+      return models;
+    } finally {
+      this.inFlightRefresh = null;
+    }
+  }
+
+  /**
+   * Clear the cache file
+   */
+  async clearCache(): Promise<void> {
+    logger.info('[clearCache] Clearing cache...');
+
+    try {
+      await secureFs.unlink(this.cacheFilePath);
+      logger.info('[clearCache] Cache cleared');
+    } catch (error) {
+      if ((error as NodeJS.ErrnoException).code !== 'ENOENT') {
+        logger.error('[clearCache] Failed to clear cache:', error);
+      }
+    }
+  }
+
+  /**
+   * Internal method to perform the actual refresh
+   */
+  private async doRefresh(): Promise<CodexModel[]> {
+    try {
+      // Check if app-server is available
+      const isAvailable = await this.appServerService.isAvailable();
+      if (!isAvailable) {
+        return [];
+      }
+
+      // Fetch models from app-server
+      const response = await this.appServerService.getModels();
+      if (!response || !response.data) {
+        return [];
+      }
+
+      // Transform models to UI format
+      const models = response.data.map((model) => this.transformModel(model));
+
+      // Save to cache
+      await this.saveToCache(models);
+
+      logger.info(`[refreshModels] ✓ Fetched fresh models (${models.length} models)`);
+
+      return models;
+    } catch (error) {
+      logger.error('[doRefresh] Refresh failed:', error);
+      return [];
+    }
+  }
+
+  /**
+   * Transform app-server model to UI-compatible format
+   */
+  private transformModel(appServerModel: AppServerModel): CodexModel {
+    return {
+      id: `codex-${appServerModel.id}`, // Add 'codex-' prefix for compatibility
+      label: appServerModel.displayName,
+      description: appServerModel.description,
+      hasThinking: appServerModel.supportedReasoningEfforts.length > 0,
+      supportsVision: true, // All Codex models support vision
+      tier: this.inferTier(appServerModel.id),
+      isDefault: appServerModel.isDefault,
+    };
+  }
+
+  /**
+   * Infer tier from model ID
+   */
+  private inferTier(modelId: string): 'premium' | 'standard' | 'basic' {
+    if (modelId.includes('max') || modelId.includes('gpt-5.2-codex')) {
+      return 'premium';
+    }
+    if (modelId.includes('mini')) {
+      return 'basic';
+    }
+    return 'standard';
+  }
+
+  /**
+   * Load cache from disk
+   */
+  private async loadFromCache(): Promise<CodexModelCache | null> {
+    try {
+      const content = await secureFs.readFile(this.cacheFilePath, 'utf-8');
+      const cache = JSON.parse(content.toString()) as CodexModelCache;
+
+      // Validate cache structure
+      if (!Array.isArray(cache.models) || typeof cache.cachedAt !== 'number') {
+        logger.warn('[loadFromCache] Invalid cache structure, ignoring');
+        return null;
+      }
+
+      return cache;
+    } catch (error) {
+      if ((error as NodeJS.ErrnoException).code !== 'ENOENT') {
+        logger.warn('[loadFromCache] Failed to read cache:', error);
+      }
+      return null;
+    }
+  }
+
+  /**
+   * Save cache to disk (atomic write)
+   */
+  private async saveToCache(models: CodexModel[]): Promise<void> {
+    const cache: CodexModelCache = {
+      models,
+      cachedAt: Date.now(),
+      ttl: this.ttl,
+    };
+
+    const tempPath = `${this.cacheFilePath}.tmp.${Date.now()}`;
+
+    try {
+      // Write to temp file
+      const content = JSON.stringify(cache, null, 2);
+      await secureFs.writeFile(tempPath, content, 'utf-8');
+
+      // Atomic rename
+      await secureFs.rename(tempPath, this.cacheFilePath);
+    } catch (error) {
+      logger.error('[saveToCache] Failed to save cache:', error);
+
+      // Clean up temp file
+      try {
+        await secureFs.unlink(tempPath);
+      } catch {
+        // Ignore cleanup errors
+      }
+    }
+  }
+}

apps/server/src/services/codex-usage-service.ts

@@ -1,11 +1,11 @@
 import {
   findCodexCliPath,
-  spawnProcess,
   getCodexAuthPath,
   systemPathExists,
   systemPathReadFile,
 } from '@automaker/platform';
 import { createLogger } from '@automaker/utils';
+import type { CodexAppServerService } from './codex-app-server-service.js';
 
 const logger = createLogger('CodexUsage');
 
@@ -18,19 +18,12 @@ export interface CodexRateLimitWindow {
   resetsAt: number;
 }
 
-export interface CodexCreditsSnapshot {
-  balance?: string;
-  unlimited?: boolean;
-  hasCredits?: boolean;
-}
-
 export type CodexPlanType = 'free' | 'plus' | 'pro' | 'team' | 'enterprise' | 'edu' | 'unknown';
 
 export interface CodexUsageData {
   rateLimits: {
     primary?: CodexRateLimitWindow;
     secondary?: CodexRateLimitWindow;
-    credits?: CodexCreditsSnapshot;
     planType?: CodexPlanType;
   } | null;
   lastUpdated: string;
@@ -39,13 +32,24 @@ export interface CodexUsageData {
 /**
  * Codex Usage Service
  *
- * Attempts to fetch usage data from Codex CLI and OpenAI API.
- * Codex CLI doesn't provide a direct usage command, but we can:
- * 1. Parse usage info from error responses (rate limit errors contain plan info)
- * 2. Check for OpenAI API usage if API key is available
+ * Fetches usage data from Codex CLI using the app-server JSON-RPC API.
+ * Falls back to auth file parsing if app-server is unavailable.
  */
 export class CodexUsageService {
   private cachedCliPath: string | null = null;
+  private appServerService: CodexAppServerService | null = null;
+  private accountPlanTypeArray: CodexPlanType[] = [
+    'free',
+    'plus',
+    'pro',
+    'team',
+    'enterprise',
+    'edu',
+  ];
+
+  constructor(appServerService?: CodexAppServerService) {
+    this.appServerService = appServerService || null;
+  }
 
   /**
    * Check if Codex CLI is available on the system
@@ -58,60 +62,131 @@ export class CodexUsageService {
   /**
    * Attempt to fetch usage data
    *
-   * Tries multiple approaches:
-   * 1. Always try to get plan type from auth file first (authoritative source)
-   * 2. Check for OpenAI API key in environment for API usage
-   * 3. Make a test request to capture rate limit headers from CLI
-   * 4. Combine results from auth file and CLI
+   * Priority order:
+   * 1. Codex app-server JSON-RPC API (most reliable, provides real-time data)
+   * 2. Auth file JWT parsing (fallback for plan type)
    */
   async fetchUsageData(): Promise<CodexUsageData> {
+    logger.info('[fetchUsageData] Starting...');
     const cliPath = this.cachedCliPath || (await findCodexCliPath());
 
     if (!cliPath) {
+      logger.error('[fetchUsageData] Codex CLI not found');
       throw new Error('Codex CLI not found. Please install it with: npm install -g @openai/codex');
     }
 
-    // Always try to get plan type from auth file first - this is the authoritative source
-    const authPlanType = await this.getPlanTypeFromAuthFile();
+    logger.info(`[fetchUsageData] Using CLI path: ${cliPath}`);
 
-    // Check if user has an API key that we can use
-    const hasApiKey = !!process.env.OPENAI_API_KEY;
-
-    if (hasApiKey) {
-      // Try to get usage from OpenAI API
-      const openaiUsage = await this.fetchOpenAIUsage();
-      if (openaiUsage) {
-        // Merge with auth file plan type if available
-        if (authPlanType && openaiUsage.rateLimits) {
-          openaiUsage.rateLimits.planType = authPlanType;
-        }
-        return openaiUsage;
-      }
+    // Try to get usage from Codex app-server (most reliable method)
+    const appServerUsage = await this.fetchFromAppServer();
+    if (appServerUsage) {
+      logger.info('[fetchUsageData] ✓ Fetched usage from app-server');
+      return appServerUsage;
     }
 
-    // Try to get usage from Codex CLI by making a simple request
-    const codexUsage = await this.fetchCodexUsage(cliPath, authPlanType);
-    if (codexUsage) {
-      return codexUsage;
-    }
+    logger.info('[fetchUsageData] App-server failed, trying auth file fallback...');
 
-    // Fallback: try to parse full usage from auth file
+    // Fallback: try to parse usage from auth file
     const authUsage = await this.fetchFromAuthFile();
     if (authUsage) {
+      logger.info('[fetchUsageData] ✓ Fetched usage from auth file');
       return authUsage;
     }
 
-    // If all else fails, return a message with helpful information
-    throw new Error(
-      'Codex usage statistics require additional configuration. ' +
-        'To enable usage tracking:\n\n' +
-        '1. Set your OpenAI API key in the environment:\n' +
-        '   export OPENAI_API_KEY=sk-...\n\n' +
-        '2. Or check your usage at:\n' +
-        '   https://platform.openai.com/usage\n\n' +
-        'Note: If using Codex CLI with ChatGPT OAuth authentication, ' +
-        'usage data must be queried through your OpenAI account.'
-    );
+    logger.info('[fetchUsageData] All methods failed, returning unknown');
+
+    // If all else fails, return unknown
+    return {
+      rateLimits: {
+        planType: 'unknown',
+      },
+      lastUpdated: new Date().toISOString(),
+    };
+  }
+
+  /**
+   * Fetch usage data from Codex app-server using JSON-RPC API
+   * This is the most reliable method as it gets real-time data from OpenAI
+   */
+  private async fetchFromAppServer(): Promise<CodexUsageData | null> {
+    try {
+      // Use CodexAppServerService if available
+      if (!this.appServerService) {
+        return null;
+      }
+
+      // Fetch account and rate limits in parallel
+      const [accountResult, rateLimitsResult] = await Promise.all([
+        this.appServerService.getAccount(),
+        this.appServerService.getRateLimits(),
+      ]);
+
+      if (!accountResult) {
+        return null;
+      }
+
+      // Build response
+      // Prefer planType from rateLimits (more accurate/current) over account (can be stale)
+      let planType: CodexPlanType = 'unknown';
+
+      // First try rate limits planType (most accurate)
+      const rateLimitsPlanType = rateLimitsResult?.rateLimits?.planType;
+      if (rateLimitsPlanType) {
+        const normalizedType = rateLimitsPlanType.toLowerCase() as CodexPlanType;
+        if (this.accountPlanTypeArray.includes(normalizedType)) {
+          planType = normalizedType;
+        }
+      }
+
+      // Fall back to account planType if rate limits didn't have it
+      if (planType === 'unknown' && accountResult.account?.planType) {
+        const normalizedType = accountResult.account.planType.toLowerCase() as CodexPlanType;
+        if (this.accountPlanTypeArray.includes(normalizedType)) {
+          planType = normalizedType;
+        }
+      }
+
+      const result: CodexUsageData = {
+        rateLimits: {
+          planType,
+        },
+        lastUpdated: new Date().toISOString(),
+      };
+
+      // Add rate limit info if available
+      if (rateLimitsResult?.rateLimits?.primary) {
+        const primary = rateLimitsResult.rateLimits.primary;
+        result.rateLimits!.primary = {
+          limit: -1, // Not provided by API
+          used: -1, // Not provided by API
+          remaining: -1, // Not provided by API
+          usedPercent: primary.usedPercent,
+          windowDurationMins: primary.windowDurationMins,
+          resetsAt: primary.resetsAt,
+        };
+      }
+
+      // Add secondary rate limit if available
+      if (rateLimitsResult?.rateLimits?.secondary) {
+        const secondary = rateLimitsResult.rateLimits.secondary;
+        result.rateLimits!.secondary = {
+          limit: -1, // Not provided by API
+          used: -1, // Not provided by API
+          remaining: -1, // Not provided by API
+          usedPercent: secondary.usedPercent,
+          windowDurationMins: secondary.windowDurationMins,
+          resetsAt: secondary.resetsAt,
+        };
+      }
+
+      logger.info(
+        `[fetchFromAppServer] ✓ Plan: ${planType}, Primary: ${result.rateLimits?.primary?.usedPercent || 'N/A'}%, Secondary: ${result.rateLimits?.secondary?.usedPercent || 'N/A'}%`
+      );
+      return result;
+    } catch (error) {
+      logger.error('[fetchFromAppServer] Failed:', error);
+      return null;
+    }
   }
 
   /**
@@ -121,9 +196,11 @@ export class CodexUsageService {
   private async getPlanTypeFromAuthFile(): Promise<CodexPlanType> {
     try {
       const authFilePath = getCodexAuthPath();
-      const exists = await systemPathExists(authFilePath);
+      logger.info(`[getPlanTypeFromAuthFile] Auth file path: ${authFilePath}`);
+      const exists = systemPathExists(authFilePath);
 
       if (!exists) {
+        logger.warn('[getPlanTypeFromAuthFile] Auth file does not exist');
         return 'unknown';
       }
 
@@ -131,16 +208,24 @@ export class CodexUsageService {
       const authData = JSON.parse(authContent);
 
       if (!authData.tokens?.id_token) {
+        logger.info('[getPlanTypeFromAuthFile] No id_token in auth file');
         return 'unknown';
       }
 
       const claims = this.parseJwt(authData.tokens.id_token);
       if (!claims) {
+        logger.info('[getPlanTypeFromAuthFile] Failed to parse JWT');
         return 'unknown';
       }
 
+      logger.info('[getPlanTypeFromAuthFile] JWT claims keys:', Object.keys(claims));
+
       // Extract plan type from nested OpenAI auth object with type validation
       const openaiAuthClaim = claims['https://api.openai.com/auth'];
+      logger.info(
+        '[getPlanTypeFromAuthFile] OpenAI auth claim:',
+        JSON.stringify(openaiAuthClaim, null, 2)
+      );
 
       let accountType: string | undefined;
       let isSubscriptionExpired = false;
@@ -188,154 +273,23 @@ export class CodexUsageService {
       }
 
       if (accountType) {
-        const normalizedType = accountType.toLowerCase();
-        if (['free', 'plus', 'pro', 'team', 'enterprise', 'edu'].includes(normalizedType)) {
-          return normalizedType as CodexPlanType;
-        }
-      }
-    } catch (error) {
-      logger.error('Failed to get plan type from auth file:', error);
-    }
-
-    return 'unknown';
-  }
-
-  /**
-   * Try to fetch usage from OpenAI API using the API key
-   */
-  private async fetchOpenAIUsage(): Promise<CodexUsageData | null> {
-    const apiKey = process.env.OPENAI_API_KEY;
-    if (!apiKey) {
-      return null;
-    }
-
-    try {
-      const endTime = Math.floor(Date.now() / 1000);
-      const startTime = endTime - 7 * 24 * 60 * 60; // Last 7 days
-
-      const response = await fetch(
-        `https://api.openai.com/v1/organization/usage/completions?start_time=${startTime}&end_time=${endTime}&limit=1`,
-        {
-          headers: {
-            Authorization: `Bearer ${apiKey}`,
-            'Content-Type': 'application/json',
-          },
-        }
-      );
-
-      if (response.ok) {
-        const data = await response.json();
-        return this.parseOpenAIUsage(data);
-      }
-    } catch (error) {
-      logger.error('Failed to fetch from OpenAI API:', error);
-    }
-
-    return null;
-  }
-
-  /**
-   * Parse OpenAI usage API response
-   */
-  private parseOpenAIUsage(data: any): CodexUsageData {
-    let totalInputTokens = 0;
-    let totalOutputTokens = 0;
-
-    if (data.data && Array.isArray(data.data)) {
-      for (const bucket of data.data) {
-        if (bucket.results && Array.isArray(bucket.results)) {
-          for (const result of bucket.results) {
-            totalInputTokens += result.input_tokens || 0;
-            totalOutputTokens += result.output_tokens || 0;
-          }
-        }
-      }
-    }
-
-    return {
-      rateLimits: {
-        planType: 'unknown',
-        credits: {
-          hasCredits: true,
-        },
-      },
-      lastUpdated: new Date().toISOString(),
-    };
-  }
-
-  /**
-   * Try to fetch usage by making a test request to Codex CLI
-   * and parsing rate limit information from the response
-   */
-  private async fetchCodexUsage(
-    cliPath: string,
-    authPlanType: CodexPlanType
-  ): Promise<CodexUsageData | null> {
-    try {
-      // Make a simple request to trigger rate limit info if at limit
-      const result = await spawnProcess({
-        command: cliPath,
-        args: ['exec', '--', 'echo', 'test'],
-        cwd: process.cwd(),
-        env: {
-          ...process.env,
-          TERM: 'dumb',
-        },
-        timeout: 10000,
-      });
-
-      // Parse the output for rate limit information
-      const combinedOutput = (result.stdout + result.stderr).toLowerCase();
-
-      // Check if we got a rate limit error
-      const rateLimitMatch = combinedOutput.match(
-        /usage_limit_reached.*?"plan_type":"([^"]+)".*?"resets_at":(\d+).*?"resets_in_seconds":(\d+)/
-      );
-
-      if (rateLimitMatch) {
-        // Rate limit error contains the plan type - use that as it's the most authoritative
-        const planType = rateLimitMatch[1] as CodexPlanType;
-        const resetsAt = parseInt(rateLimitMatch[2], 10);
-        const resetsInSeconds = parseInt(rateLimitMatch[3], 10);
-
+        const normalizedType = accountType.toLowerCase() as CodexPlanType;
         logger.info(
-          `Rate limit hit - plan: ${planType}, resets in ${Math.ceil(resetsInSeconds / 60)} mins`
+          `[getPlanTypeFromAuthFile] Account type: "${accountType}", normalized: "${normalizedType}"`
         );
-
-        return {
-          rateLimits: {
-            planType,
-            primary: {
-              limit: 0,
-              used: 0,
-              remaining: 0,
-              usedPercent: 100,
-              windowDurationMins: Math.ceil(resetsInSeconds / 60),
-              resetsAt,
-            },
-          },
-          lastUpdated: new Date().toISOString(),
-        };
+        if (this.accountPlanTypeArray.includes(normalizedType)) {
+          logger.info(`[getPlanTypeFromAuthFile] Returning plan type: ${normalizedType}`);
+          return normalizedType;
+        }
+      } else {
+        logger.info('[getPlanTypeFromAuthFile] No account type found in claims');
       }
-
-      // No rate limit error - use the plan type from auth file
-      const isFreePlan = authPlanType === 'free';
-
-      return {
-        rateLimits: {
-          planType: authPlanType,
-          credits: {
-            hasCredits: true,
-            unlimited: !isFreePlan && authPlanType !== 'unknown',
-          },
-        },
-        lastUpdated: new Date().toISOString(),
-      };
     } catch (error) {
-      logger.error('Failed to fetch from Codex CLI:', error);
+      logger.error('[getPlanTypeFromAuthFile] Failed to get plan type from auth file:', error);
     }
 
-    return null;
+    logger.info('[getPlanTypeFromAuthFile] Returning unknown');
+    return 'unknown';
   }
 
   /**
@@ -343,27 +297,27 @@ export class CodexUsageService {
    * Reuses getPlanTypeFromAuthFile to avoid code duplication
    */
   private async fetchFromAuthFile(): Promise<CodexUsageData | null> {
+    logger.info('[fetchFromAuthFile] Starting...');
     try {
       const planType = await this.getPlanTypeFromAuthFile();
+      logger.info(`[fetchFromAuthFile] Got plan type: ${planType}`);
 
       if (planType === 'unknown') {
+        logger.info('[fetchFromAuthFile] Plan type unknown, returning null');
         return null;
       }
 
-      const isFreePlan = planType === 'free';
-
-      return {
+      const result: CodexUsageData = {
         rateLimits: {
           planType,
-          credits: {
-            hasCredits: true,
-            unlimited: !isFreePlan,
-          },
         },
         lastUpdated: new Date().toISOString(),
       };
+
+      logger.info('[fetchFromAuthFile] Returning result:', JSON.stringify(result, null, 2));
+      return result;
     } catch (error) {
-      logger.error('Failed to parse auth file:', error);
+      logger.error('[fetchFromAuthFile] Failed to parse auth file:', error);
     }
 
     return null;
@@ -372,7 +326,7 @@ export class CodexUsageService {
   /**
    * Parse JWT token to extract claims
    */
-  private parseJwt(token: string): any {
+  private parseJwt(token: string): Record<string, unknown> | null {
     try {
       const parts = token.split('.');
 
@@ -383,18 +337,8 @@ export class CodexUsageService {
       const base64Url = parts[1];
       const base64 = base64Url.replace(/-/g, '+').replace(/_/g, '/');
 
-      // Use Buffer for Node.js environment instead of atob
-      let jsonPayload: string;
-      if (typeof Buffer !== 'undefined') {
-        jsonPayload = Buffer.from(base64, 'base64').toString('utf-8');
-      } else {
-        jsonPayload = decodeURIComponent(
-          atob(base64)
-            .split('')
-            .map((c) => '%' + ('00' + c.charCodeAt(0).toString(16)).slice(-2))
-            .join('')
-        );
-      }
+      // Use Buffer for Node.js environment
+      const jsonPayload = Buffer.from(base64, 'base64').toString('utf-8');
 
       return JSON.parse(jsonPayload);
     } catch {

apps/server/src/services/feature-loader.ts

@@ -308,13 +308,15 @@ export class FeatureLoader {
    * @param updates - Partial feature updates
    * @param descriptionHistorySource - Source of description change ('enhance' or 'edit')
    * @param enhancementMode - Enhancement mode if source is 'enhance'
+   * @param preEnhancementDescription - Description before enhancement (for restoring original)
    */
   async update(
     projectPath: string,
     featureId: string,
     updates: Partial<Feature>,
     descriptionHistorySource?: 'enhance' | 'edit',
-    enhancementMode?: 'improve' | 'technical' | 'simplify' | 'acceptance'
+    enhancementMode?: 'improve' | 'technical' | 'simplify' | 'acceptance' | 'ux-reviewer',
+    preEnhancementDescription?: string
   ): Promise<Feature> {
     const feature = await this.get(projectPath, featureId);
     if (!feature) {
@@ -338,9 +340,31 @@ export class FeatureLoader {
       updates.description !== feature.description &&
       updates.description.trim()
     ) {
+      const timestamp = new Date().toISOString();
+
+      // If this is an enhancement and we have the pre-enhancement description,
+      // add the original text to history first (so user can restore to it)
+      if (
+        descriptionHistorySource === 'enhance' &&
+        preEnhancementDescription &&
+        preEnhancementDescription.trim()
+      ) {
+        // Check if this pre-enhancement text is different from the last history entry
+        const lastEntry = updatedHistory[updatedHistory.length - 1];
+        if (!lastEntry || lastEntry.description !== preEnhancementDescription) {
+          const preEnhanceEntry: DescriptionHistoryEntry = {
+            description: preEnhancementDescription,
+            timestamp,
+            source: updatedHistory.length === 0 ? 'initial' : 'edit',
+          };
+          updatedHistory = [...updatedHistory, preEnhanceEntry];
+        }
+      }
+
+      // Add the new/enhanced description to history
       const historyEntry: DescriptionHistoryEntry = {
         description: updates.description,
-        timestamp: new Date().toISOString(),
+        timestamp,
         source: descriptionHistorySource || 'edit',
         ...(descriptionHistorySource === 'enhance' && enhancementMode ? { enhancementMode } : {}),
       };

apps/server/src/services/settings-service.ts

@@ -22,7 +22,6 @@ import type {
   Credentials,
   ProjectSettings,
   KeyboardShortcuts,
-  AIProfile,
   ProjectRef,
   TrashedProjectRef,
   BoardBackgroundSettings,
@@ -299,7 +298,6 @@ export class SettingsService {
     ignoreEmptyArrayOverwrite('trashedProjects');
     ignoreEmptyArrayOverwrite('projectHistory');
     ignoreEmptyArrayOverwrite('recentFolders');
-    ignoreEmptyArrayOverwrite('aiProfiles');
     ignoreEmptyArrayOverwrite('mcpServers');
     ignoreEmptyArrayOverwrite('enabledCursorModels');
 
@@ -602,8 +600,6 @@ export class SettingsService {
         theme: (appState.theme as GlobalSettings['theme']) || 'dark',
         sidebarOpen: appState.sidebarOpen !== undefined ? (appState.sidebarOpen as boolean) : true,
         chatHistoryOpen: (appState.chatHistoryOpen as boolean) || false,
-        kanbanCardDetailLevel:
-          (appState.kanbanCardDetailLevel as GlobalSettings['kanbanCardDetailLevel']) || 'standard',
         maxConcurrency: (appState.maxConcurrency as number) || 3,
         defaultSkipTests:
           appState.defaultSkipTests !== undefined ? (appState.defaultSkipTests as boolean) : true,
@@ -617,18 +613,15 @@ export class SettingsService {
             : false,
         useWorktrees:
           appState.useWorktrees !== undefined ? (appState.useWorktrees as boolean) : true,
-        showProfilesOnly: (appState.showProfilesOnly as boolean) || false,
         defaultPlanningMode:
           (appState.defaultPlanningMode as GlobalSettings['defaultPlanningMode']) || 'skip',
         defaultRequirePlanApproval: (appState.defaultRequirePlanApproval as boolean) || false,
-        defaultAIProfileId: (appState.defaultAIProfileId as string | null) || null,
         muteDoneSound: (appState.muteDoneSound as boolean) || false,
         enhancementModel:
           (appState.enhancementModel as GlobalSettings['enhancementModel']) || 'sonnet',
         keyboardShortcuts:
           (appState.keyboardShortcuts as KeyboardShortcuts) ||
           DEFAULT_GLOBAL_SETTINGS.keyboardShortcuts,
-        aiProfiles: (appState.aiProfiles as AIProfile[]) || [],
         projects: (appState.projects as ProjectRef[]) || [],
         trashedProjects: (appState.trashedProjects as TrashedProjectRef[]) || [],
         projectHistory: (appState.projectHistory as string[]) || [],

apps/server/src/types/settings.ts

@@ -7,13 +7,11 @@
 
 export type {
   ThemeMode,
-  KanbanCardDetailLevel,
   ModelAlias,
   PlanningMode,
   ThinkingLevel,
   ModelProvider,
   KeyboardShortcuts,
-  AIProfile,
   ProjectRef,
   TrashedProjectRef,
   ChatSessionRef,