chore: add release-check CI to check if we are in release mode

* chore: improve pre-release CI to just use dropdown

* chore: implement requested changes

.changeset/eleven-horses-shop.md

@@ -1,5 +0,0 @@
----
-"task-master-ai": patch
----
-
-Fix for tasks not found when using string IDs

.changeset/fix-tag-complexity-detection.md

@@ -1,7 +0,0 @@
----
-"task-master-ai": patch
----
-
-Fix tag-specific complexity report detection in expand command
-
-The expand command now correctly finds and uses tag-specific complexity reports (e.g., `task-complexity-report_feature-xyz.json`) when operating in a tag context. Previously, it would always look for the generic `task-complexity-report.json` file due to a default value in the CLI option definition.

.changeset/floppy-news-buy-1.md

@@ -1,38 +0,0 @@
----
-"task-master-ai": minor
----
-
-Add new `scope-up` and `scope-down` commands for dynamic task complexity adjustment
-
-This release introduces two powerful new commands that allow you to dynamically adjust the complexity of your tasks and subtasks without recreating them from scratch.
-
-**New CLI Commands:**
-- `task-master scope-up` - Increase task complexity (add more detail, requirements, or implementation steps)
-- `task-master scope-down` - Decrease task complexity (simplify, remove unnecessary details, or streamline)
-
-**Key Features:**
-- **Multiple tasks**: Support comma-separated IDs to adjust multiple tasks at once (`--id=5,7,12`)
-- **Strength levels**: Choose adjustment intensity with `--strength=light|regular|heavy` (defaults to regular)
-- **Custom prompts**: Use `--prompt` flag to specify exactly how you want tasks adjusted
-- **MCP integration**: Available as `scope_up_task` and `scope_down_task` tools in Cursor and other MCP environments
-- **Smart context**: AI considers your project context and task dependencies when making adjustments
-
-**Usage Examples:**
-```bash
-# Make a task more detailed
-task-master scope-up --id=5
-
-# Simplify multiple tasks with light touch
-task-master scope-down --id=10,11,12 --strength=light
-
-# Custom adjustment with specific instructions  
-task-master scope-up --id=7 --prompt="Add more error handling and edge cases"
-```
-
-**Why use this?**
-- **Iterative refinement**: Adjust task complexity as your understanding evolves
-- **Project phase adaptation**: Scale tasks up for implementation, down for planning
-- **Team coordination**: Adjust complexity based on team member experience levels
-- **Milestone alignment**: Fine-tune tasks to match project phase requirements
-
-Perfect for agile workflows where task requirements change as you learn more about the problem space.

.changeset/fuzzy-words-count.md

@@ -0,0 +1,8 @@
+---
+"task-master-ai": patch
+---
+
+Fix scope-up/down prompts to include all required fields for better AI model compatibility
+
+- Added missing `priority` field to scope adjustment prompts to prevent validation errors with Claude-code and other models
+- Ensures generated JSON includes all fields required by the schema

.changeset/pre.json

@@ -1,14 +1,13 @@
 {
-  "mode": "pre",
+  "mode": "exit",
   "tag": "rc",
   "initialVersions": {
-    "task-master-ai": "0.22.0",
+    "task-master-ai": "0.23.0",
-    "extension": "0.20.0"
+    "extension": "0.23.0"
   },
   "changesets": [
-    "eleven-horses-shop",
+    "fuzzy-words-count",
-    "fix-tag-complexity-detection",
+    "tender-trams-refuse",
-    "floppy-news-buy-1",
+    "vast-sites-leave"
-    "sour-pans-beam-1"
   ]
 }

.changeset/sour-pans-beam-1.md

@@ -1,42 +0,0 @@
----
-"extension": minor
----
-
-🎉 **Introducing TaskMaster Extension!** 
-
-We're thrilled to launch the first version of our Code extension, bringing the power of TaskMaster directly into your favorite code editor. While this is our initial release and we've kept things focused, it already packs powerful features to supercharge your development workflow.
-
-## ✨ Key Features
-
-### 📋 Visual Task Management
-- **Kanban Board View**: Visualize all your tasks in an intuitive board layout directly in VS Code
-- **Drag & Drop**: Easily change task status by dragging cards between columns
-- **Real-time Updates**: See changes instantly as you work through your project
-
-### 🏷️ Multi-Context Support
-- **Tag Switching**: Seamlessly switch between different project contexts/tags
-- **Isolated Workflows**: Keep different features or experiments organized separately
-
-### 🤖 AI-Powered Task Updates
-- **Smart Updates**: Use TaskMaster's AI capabilities to update tasks and subtasks
-- **Context-Aware**: Leverages your existing TaskMaster configuration and models
-
-### 📊 Rich Task Information
-- **Complexity Scores**: See task complexity ratings at a glance
-- **Subtask Visualization**: Expand tasks to view and manage subtasks
-- **Dependency Graphs**: Understand task relationships and dependencies visually
-
-### ⚙️ Configuration Management
-- **Visual Config Editor**: View and understand your `.taskmaster/config.json` settings
-- **Easy Access**: No more manual JSON editing for common configuration tasks
-
-### 🚀 Quick Actions
-- **Status Updates**: Change task status with a single click
-- **Task Details**: Access full task information without leaving VS Code
-- **Integrated Commands**: All TaskMaster commands available through the command palette
-
-## 🎯 What's Next?
-
-This is just the beginning! We wanted to get a solid foundation into your hands quickly. The extension will evolve rapidly with your feedback, adding more advanced features, better visualizations, and deeper integration with your development workflow.
-
-Thank you for being part of the TaskMaster journey. Your workflow has never looked better! 🚀

.changeset/tender-trams-refuse.md

@@ -0,0 +1,8 @@
+---
+"task-master-ai": patch
+---
+
+Fix MCP scope-up/down tools not finding tasks
+
+- Fixed task ID parsing in MCP layer - now correctly converts string IDs to numbers
+- scope_up_task and scope_down_task MCP tools now work properly

.changeset/vast-sites-leave.md

@@ -0,0 +1,11 @@
+---
+"task-master-ai": patch
+---
+
+Improve AI provider compatibility for JSON generation
+
+- Fixed schema compatibility issues between Perplexity and OpenAI o3 models
+- Removed nullable/default modifiers from Zod schemas for broader compatibility
+- Added automatic JSON repair for malformed AI responses (handles cases like missing array values)
+- Perplexity now uses JSON mode for more reliable structured output
+- Post-processing handles default values separately from schema validation

.github/scripts/tag-extension.mjs

@@ -1,15 +1,37 @@
 #!/usr/bin/env node
 import assert from 'node:assert/strict';
 import { spawnSync } from 'node:child_process';
-import { readFileSync } from 'node:fs';
+import { readFileSync, existsSync } from 'node:fs';
-import { join, dirname } from 'node:path';
+import { join, dirname, resolve } from 'node:path';
 import { fileURLToPath } from 'node:url';
 
 const __filename = fileURLToPath(import.meta.url);
 const __dirname = dirname(__filename);
 
+// Find the root directory by looking for package.json
+function findRootDir(startDir) {
+	let currentDir = resolve(startDir);
+	while (currentDir !== '/') {
+		if (existsSync(join(currentDir, 'package.json'))) {
+			// Verify it's the root package.json by checking for expected fields
+			try {
+				const pkg = JSON.parse(
+					readFileSync(join(currentDir, 'package.json'), 'utf8')
+				);
+				if (pkg.name === 'task-master-ai' || pkg.repository) {
+					return currentDir;
+				}
+			} catch {}
+		}
+		currentDir = dirname(currentDir);
+	}
+	throw new Error('Could not find root directory');
+}
+
+const rootDir = findRootDir(__dirname);
+
 // Read the extension's package.json
-const extensionDir = join(__dirname, '..', 'apps', 'extension');
+const extensionDir = join(rootDir, 'apps', 'extension');
 const pkgPath = join(extensionDir, 'package.json');
 
 let pkg;
@@ -22,7 +44,7 @@ try {
 }
 
 // Read root package.json for repository info
-const rootPkgPath = join(__dirname, '..', 'package.json');
+const rootPkgPath = join(rootDir, 'package.json');
 let rootPkg;
 try {
 	const rootPkgContent = readFileSync(rootPkgPath, 'utf8');
@@ -40,13 +62,51 @@ assert(rootPkg.repository, 'root package.json must have a repository field');
 const tag = `${pkg.name}@${pkg.version}`;
 
 // Get repository URL from root package.json
-const repoUrl = rootPkg.repository.url;
+// Get repository URL and clean it up for git ls-remote
+let repoUrl = rootPkg.repository.url || rootPkg.repository;
+if (typeof repoUrl === 'string') {
+	// Convert git+https://github.com/... to https://github.com/...
+	repoUrl = repoUrl.replace(/^git\+/, '');
+	// Ensure it ends with .git for proper remote access
+	if (!repoUrl.endsWith('.git')) {
+		repoUrl += '.git';
+	}
+}
 
-const { status, stdout, error } = spawnSync('git', ['ls-remote', repoUrl, tag]);
+console.log(`Checking remote repository: ${repoUrl} for tag: ${tag}`);
 
-assert.equal(status, 0, error);
+let gitResult = spawnSync('git', ['ls-remote', repoUrl, tag], {
+	encoding: 'utf8',
+	env: { ...process.env }
+});
 
-const exists = String(stdout).trim() !== '';
+if (gitResult.status !== 0) {
+	console.error('Git ls-remote failed:');
+	console.error('Exit code:', gitResult.status);
+	console.error('Error:', gitResult.error);
+	console.error('Stderr:', gitResult.stderr);
+	console.error('Command:', `git ls-remote ${repoUrl} ${tag}`);
+
+	// For CI environments, try using origin instead of the full URL
+	if (process.env.CI) {
+		console.log('Retrying with origin remote...');
+		gitResult = spawnSync('git', ['ls-remote', 'origin', tag], {
+			encoding: 'utf8'
+		});
+
+		if (gitResult.status !== 0) {
+			throw new Error(
+				`Failed to check remote for tag ${tag}. Exit code: ${gitResult.status}`
+			);
+		}
+	} else {
+		throw new Error(
+			`Failed to check remote for tag ${tag}. Exit code: ${gitResult.status}`
+		);
+	}
+}
+
+const exists = String(gitResult.stdout).trim() !== '';
 
 if (!exists) {
 	console.log(`Creating new extension tag: ${tag}`);

.github/workflows/pre-release.yml

@@ -3,11 +3,12 @@ name: Pre-Release (RC)
 on:
   workflow_dispatch: # Allows manual triggering from GitHub UI/API
 
-concurrency: pre-release-${{ github.ref }}
+concurrency: pre-release-${{ github.ref_name }}
-
 jobs:
   rc:
     runs-on: ubuntu-latest
+    # Only allow pre-releases on non-main branches
+    if: github.ref != 'refs/heads/main'
     steps:
       - uses: actions/checkout@v4
         with:

.github/workflows/release-check.yml

@@ -0,0 +1,36 @@
+name: Release Check
+
+on:
+  pull_request:
+    branches:
+      - main
+
+jobs:
+  check-release-mode:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Check release mode
+        run: |
+          set -euo pipefail
+          echo "🔍 Checking if branch is in pre-release mode..."
+
+          if [[ -f .changeset/pre.json ]]; then
+            echo "❌ ERROR: This branch is in pre-release mode!"
+            echo ""
+            echo "Pre-release mode must be exited before merging to main."
+            echo ""
+            echo "To fix this, run the following commands in your branch:"
+            echo "  npx changeset pre exit"
+            echo "  git add -u"
+            echo "  git commit -m 'chore: exit pre-release mode'"
+            echo "  git push"
+            echo ""
+            echo "Then update this pull request."
+            exit 1
+          fi
+
+          echo "✅ Not in pre-release mode - PR can be merged"

.github/workflows/release.yml

@@ -38,27 +38,27 @@ jobs:
         run: npm ci
         timeout-minutes: 2
 
-      - name: Exit pre-release mode and clean up
+      - name: Check pre-release mode
         run: |
-          echo "🔄 Ensuring we're not in pre-release mode for main branch..."
+          set -euo pipefail
-          
+          echo "🔍 Checking pre-release mode status..."
-          # Exit pre-release mode if we're in it
+          if [[ -f .changeset/pre.json ]]; then
-          npx changeset pre exit || echo "Not in pre-release mode"
+            echo "❌ ERROR: Main branch is in pre-release mode!"
-          
+            echo ""
-          # Remove pre.json file if it exists (belt and suspenders approach)
+            echo "Pre-release mode should only be used on feature branches, not main."
-          if [ -f .changeset/pre.json ]; then
+            echo ""
-            echo "🧹 Removing pre.json file..."
+            echo "To fix this, run the following commands locally:"
-            rm -f .changeset/pre.json
+            echo "  npx changeset pre exit"
-          fi
+            echo "  git add -u"
-          
+            echo "  git commit -m 'chore: exit pre-release mode'"
-          # Verify the file is gone
+            echo "  git push origin main"
-          if [ ! -f .changeset/pre.json ]; then
+            echo ""
-            echo "✅ pre.json successfully removed"
+            echo "Then re-run this workflow."
-          else
-            echo "❌ Failed to remove pre.json"
             exit 1
           fi
 
+          echo "✅ Not in pre-release mode - proceeding with release"
+
       - name: Create Release Pull Request or Publish to npm
         uses: changesets/action@v1
         with:

CHANGELOG.md

@@ -1,5 +1,72 @@
 # task-master-ai
 
+## 0.23.1-rc.0
+
+### Patch Changes
+
+- [#1079](https://github.com/eyaltoledano/claude-task-master/pull/1079) [`e495b2b`](https://github.com/eyaltoledano/claude-task-master/commit/e495b2b55950ee54c7d0f1817d8530e28bd79c05) Thanks [@Crunchyman-ralph](https://github.com/Crunchyman-ralph)! - Fix scope-up/down prompts to include all required fields for better AI model compatibility
+  - Added missing `priority` field to scope adjustment prompts to prevent validation errors with Claude-code and other models
+  - Ensures generated JSON includes all fields required by the schema
+
+- [#1079](https://github.com/eyaltoledano/claude-task-master/pull/1079) [`e495b2b`](https://github.com/eyaltoledano/claude-task-master/commit/e495b2b55950ee54c7d0f1817d8530e28bd79c05) Thanks [@Crunchyman-ralph](https://github.com/Crunchyman-ralph)! - Fix MCP scope-up/down tools not finding tasks
+  - Fixed task ID parsing in MCP layer - now correctly converts string IDs to numbers
+  - scope_up_task and scope_down_task MCP tools now work properly
+
+- [#1079](https://github.com/eyaltoledano/claude-task-master/pull/1079) [`e495b2b`](https://github.com/eyaltoledano/claude-task-master/commit/e495b2b55950ee54c7d0f1817d8530e28bd79c05) Thanks [@Crunchyman-ralph](https://github.com/Crunchyman-ralph)! - Improve AI provider compatibility for JSON generation
+  - Fixed schema compatibility issues between Perplexity and OpenAI o3 models
+  - Removed nullable/default modifiers from Zod schemas for broader compatibility
+  - Added automatic JSON repair for malformed AI responses (handles cases like missing array values)
+  - Perplexity now uses JSON mode for more reliable structured output
+  - Post-processing handles default values separately from schema validation
+
+## 0.23.0
+
+### Minor Changes
+
+- [#1064](https://github.com/eyaltoledano/claude-task-master/pull/1064) [`53903f1`](https://github.com/eyaltoledano/claude-task-master/commit/53903f1e8eee23ac512eb13a6d81d8cbcfe658cb) Thanks [@eyaltoledano](https://github.com/eyaltoledano)! - Add new `scope-up` and `scope-down` commands for dynamic task complexity adjustment
+
+  This release introduces two powerful new commands that allow you to dynamically adjust the complexity of your tasks and subtasks without recreating them from scratch.
+
+  **New CLI Commands:**
+  - `task-master scope-up` - Increase task complexity (add more detail, requirements, or implementation steps)
+  - `task-master scope-down` - Decrease task complexity (simplify, remove unnecessary details, or streamline)
+
+  **Key Features:**
+  - **Multiple tasks**: Support comma-separated IDs to adjust multiple tasks at once (`--id=5,7,12`)
+  - **Strength levels**: Choose adjustment intensity with `--strength=light|regular|heavy` (defaults to regular)
+  - **Custom prompts**: Use `--prompt` flag to specify exactly how you want tasks adjusted
+  - **MCP integration**: Available as `scope_up_task` and `scope_down_task` tools in Cursor and other MCP environments
+  - **Smart context**: AI considers your project context and task dependencies when making adjustments
+
+  **Usage Examples:**
+
+  ```bash
+  # Make a task more detailed
+  task-master scope-up --id=5
+
+  # Simplify multiple tasks with light touch
+  task-master scope-down --id=10,11,12 --strength=light
+
+  # Custom adjustment with specific instructions
+  task-master scope-up --id=7 --prompt="Add more error handling and edge cases"
+  ```
+
+  **Why use this?**
+  - **Iterative refinement**: Adjust task complexity as your understanding evolves
+  - **Project phase adaptation**: Scale tasks up for implementation, down for planning
+  - **Team coordination**: Adjust complexity based on team member experience levels
+  - **Milestone alignment**: Fine-tune tasks to match project phase requirements
+
+  Perfect for agile workflows where task requirements change as you learn more about the problem space.
+
+### Patch Changes
+
+- [#1063](https://github.com/eyaltoledano/claude-task-master/pull/1063) [`2ae6e7e`](https://github.com/eyaltoledano/claude-task-master/commit/2ae6e7e6be3605c3c4d353f34666e54750dba973) Thanks [@Crunchyman-ralph](https://github.com/Crunchyman-ralph)! - Fix for tasks not found when using string IDs
+
+- [#1049](https://github.com/eyaltoledano/claude-task-master/pull/1049) [`45a14c3`](https://github.com/eyaltoledano/claude-task-master/commit/45a14c323d21071c15106335e89ad1f4a20976ab) Thanks [@ben-vargas](https://github.com/ben-vargas)! - Fix tag-specific complexity report detection in expand command
+
+  The expand command now correctly finds and uses tag-specific complexity reports (e.g., `task-complexity-report_feature-xyz.json`) when operating in a tag context. Previously, it would always look for the generic `task-complexity-report.json` file due to a default value in the CLI option definition.
+
 ## 0.23.0-rc.2
 
 ### Minor Changes

apps/extension/CHANGELOG.md

@@ -1,5 +1,48 @@
 # Change Log
 
+## 0.23.0
+
+### Minor Changes
+
+- [#1064](https://github.com/eyaltoledano/claude-task-master/pull/1064) [`b82d858`](https://github.com/eyaltoledano/claude-task-master/commit/b82d858f81a1e702ad59d84d5ae8a2ca84359a83) Thanks [@Crunchyman-ralph](https://github.com/Crunchyman-ralph)! - 🎉 **Introducing TaskMaster Extension!**
+
+  We're thrilled to launch the first version of our Code extension, bringing the power of TaskMaster directly into your favorite code editor. While this is our initial release and we've kept things focused, it already packs powerful features to supercharge your development workflow.
+
+  ## ✨ Key Features
+
+  ### 📋 Visual Task Management
+  - **Kanban Board View**: Visualize all your tasks in an intuitive board layout directly in VS Code
+  - **Drag & Drop**: Easily change task status by dragging cards between columns
+  - **Real-time Updates**: See changes instantly as you work through your project
+
+  ### 🏷️ Multi-Context Support
+  - **Tag Switching**: Seamlessly switch between different project contexts/tags
+  - **Isolated Workflows**: Keep different features or experiments organized separately
+
+  ### 🤖 AI-Powered Task Updates
+  - **Smart Updates**: Use TaskMaster's AI capabilities to update tasks and subtasks
+  - **Context-Aware**: Leverages your existing TaskMaster configuration and models
+
+  ### 📊 Rich Task Information
+  - **Complexity Scores**: See task complexity ratings at a glance
+  - **Subtask Visualization**: Expand tasks to view and manage subtasks
+  - **Dependency Graphs**: Understand task relationships and dependencies visually
+
+  ### ⚙️ Configuration Management
+  - **Visual Config Editor**: View and understand your `.taskmaster/config.json` settings
+  - **Easy Access**: No more manual JSON editing for common configuration tasks
+
+  ### 🚀 Quick Actions
+  - **Status Updates**: Change task status with a single click
+  - **Task Details**: Access full task information without leaving VS Code
+  - **Integrated Commands**: All TaskMaster commands available through the command palette
+
+  ## 🎯 What's Next?
+
+  This is just the beginning! We wanted to get a solid foundation into your hands quickly. The extension will evolve rapidly with your feedback, adding more advanced features, better visualizations, and deeper integration with your development workflow.
+
+  Thank you for being part of the TaskMaster journey. Your workflow has never looked better! 🚀
+
 ## 0.23.0-rc.1
 
 ### Minor Changes

apps/extension/package.json

@@ -3,23 +3,15 @@
 	"private": true,
 	"displayName": "TaskMaster",
 	"description": "A visual Kanban board interface for TaskMaster projects in VS Code",
-	"version": "0.23.0-rc.1",
+	"version": "0.23.0",
 	"publisher": "Hamster",
 	"icon": "assets/icon.png",
 	"engines": {
 		"vscode": "^1.93.0"
 	},
-	"categories": [
+	"categories": ["AI", "Visualization", "Education", "Other"],
-		"AI",
-		"Visualization",
-		"Education",
-		"Other"
-	],
 	"main": "./dist/extension.js",
-	"activationEvents": [
+	"activationEvents": ["onStartupFinished", "workspaceContains:.taskmaster/**"],
-		"onStartupFinished",
-		"workspaceContains:.taskmaster/**"
-	],
 	"contributes": {
 		"viewsContainers": {
 			"activitybar": [
@@ -80,9 +72,7 @@
 					"items": {
 						"type": "string"
 					},
-					"default": [
+					"default": ["task-master-ai"],
-						"task-master-ai"
-					],
 					"description": "An array of arguments to pass to the MCP server command."
 				},
 				"taskmaster.mcp.cwd": {
@@ -149,11 +139,7 @@
 				},
 				"taskmaster.ui.theme": {
 					"type": "string",
-					"enum": [
+					"enum": ["auto", "light", "dark"],
-						"auto",
-						"light",
-						"dark"
-					],
 					"default": "auto",
 					"description": "UI theme preference"
 				},
@@ -214,12 +200,7 @@
 				},
 				"taskmaster.debug.logLevel": {
 					"type": "string",
-					"enum": [
+					"enum": ["error", "warn", "info", "debug"],
-						"error",
-						"warn",
-						"info",
-						"debug"
-					],
 					"default": "info",
 					"description": "Logging level"
 				},

mcp-server/src/core/direct-functions/scope-down.js

@@ -71,8 +71,8 @@ export async function scopeDownDirect(args, log, context = {}) {
 			};
 		}
 
-		// Parse task IDs
+		// Parse task IDs - convert to numbers as expected by scopeDownTask
-		const taskIds = id.split(',').map((taskId) => taskId.trim());
+		const taskIds = id.split(',').map((taskId) => parseInt(taskId.trim(), 10));
 
 		log.info(
 			`Scoping down tasks: ${taskIds.join(', ')}, strength: ${strength}, research: ${research}`
@@ -90,10 +90,10 @@ export async function scopeDownDirect(args, log, context = {}) {
 				projectRoot,
 				commandName: 'scope-down',
 				outputType: 'mcp',
-				tag
+				tag,
+				research
 			},
-			'json', // outputFormat
+			'json' // outputFormat
-			research
 		);
 
 		// Restore normal logging

mcp-server/src/core/direct-functions/scope-up.js

@@ -71,8 +71,8 @@ export async function scopeUpDirect(args, log, context = {}) {
 			};
 		}
 
-		// Parse task IDs
+		// Parse task IDs - convert to numbers as expected by scopeUpTask
-		const taskIds = id.split(',').map((taskId) => taskId.trim());
+		const taskIds = id.split(',').map((taskId) => parseInt(taskId.trim(), 10));
 
 		log.info(
 			`Scoping up tasks: ${taskIds.join(', ')}, strength: ${strength}, research: ${research}`
@@ -90,10 +90,10 @@ export async function scopeUpDirect(args, log, context = {}) {
 				projectRoot,
 				commandName: 'scope-up',
 				outputType: 'mcp',
-				tag
+				tag,
+				research
 			},
-			'json', // outputFormat
+			'json' // outputFormat
-			research
 		);
 
 		// Restore normal logging

package-lock.json

@@ -1,12 +1,12 @@
 {
 	"name": "task-master-ai",
-	"version": "0.22.1-rc.0",
+	"version": "0.23.0",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "task-master-ai",
-			"version": "0.22.1-rc.0",
+			"version": "0.23.0",
 			"license": "MIT WITH Commons-Clause",
 			"workspaces": [
 				"apps/*",
@@ -46,6 +46,7 @@
 				"helmet": "^8.1.0",
 				"inquirer": "^12.5.0",
 				"jsonc-parser": "^3.3.1",
+				"jsonrepair": "^3.13.0",
 				"jsonwebtoken": "^9.0.2",
 				"lru-cache": "^10.2.0",
 				"ollama-ai-provider": "^1.2.0",
@@ -84,7 +85,7 @@
 			}
 		},
 		"apps/extension": {
-			"version": "0.22.3",
+			"version": "0.23.0",
 			"devDependencies": {
 				"@dnd-kit/core": "^6.3.1",
 				"@dnd-kit/modifiers": "^9.0.0",
@@ -14942,6 +14943,15 @@
 				"graceful-fs": "^4.1.6"
 			}
 		},
+		"node_modules/jsonrepair": {
+			"version": "3.13.0",
+			"resolved": "https://registry.npmjs.org/jsonrepair/-/jsonrepair-3.13.0.tgz",
+			"integrity": "sha512-5YRzlAQ7tuzV1nAJu3LvDlrKtBFIALHN2+a+I1MGJCt3ldRDBF/bZuvIPzae8Epot6KBXd0awRZZcuoeAsZ/mw==",
+			"license": "ISC",
+			"bin": {
+				"jsonrepair": "bin/cli.js"
+			}
+		},
 		"node_modules/jsonwebtoken": {
 			"version": "9.0.2",
 			"resolved": "https://registry.npmjs.org/jsonwebtoken/-/jsonwebtoken-9.0.2.tgz",

package.json

@@ -1,6 +1,6 @@
 {
 	"name": "task-master-ai",
-	"version": "0.23.0-rc.2",
+	"version": "0.23.1-rc.0",
 	"description": "A task management system for ambitious AI-driven development that doesn't overwhelm and confuse Cursor.",
 	"main": "index.js",
 	"type": "module",
@@ -9,10 +9,7 @@
 		"task-master-mcp": "mcp-server/server.js",
 		"task-master-ai": "mcp-server/server.js"
 	},
-	"workspaces": [
+	"workspaces": ["apps/*", "."],
-		"apps/*",
-		"."
-	],
 	"scripts": {
 		"test": "node --experimental-vm-modules node_modules/.bin/jest",
 		"test:fails": "node --experimental-vm-modules node_modules/.bin/jest --onlyFailures",
@@ -76,6 +73,7 @@
 		"helmet": "^8.1.0",
 		"inquirer": "^12.5.0",
 		"jsonc-parser": "^3.3.1",
+		"jsonrepair": "^3.13.0",
 		"jsonwebtoken": "^9.0.2",
 		"lru-cache": "^10.2.0",
 		"ollama-ai-provider": "^1.2.0",

scripts/modules/commands.js

@@ -1479,7 +1479,8 @@ function registerCommands(programInstance) {
 					projectRoot: taskMaster.getProjectRoot(),
 					tag,
 					commandName: 'scope-up',
-					outputType: 'cli'
+					outputType: 'cli',
+					research: options.research || false
 				};
 
 				const result = await scopeUpTask(
@@ -1605,7 +1606,8 @@ function registerCommands(programInstance) {
 					projectRoot: taskMaster.getProjectRoot(),
 					tag,
 					commandName: 'scope-down',
-					outputType: 'cli'
+					outputType: 'cli',
+					research: options.research || false
 				};
 
 				const result = await scopeDownTask(

scripts/modules/task-manager/parse-prd.js

@@ -23,14 +23,14 @@ import { displayAiUsageSummary } from '../ui.js';
 
 // Define the Zod schema for a SINGLE task object
 const prdSingleTaskSchema = z.object({
-	id: z.number().int().positive(),
+	id: z.number(),
 	title: z.string().min(1),
 	description: z.string().min(1),
-	details: z.string().nullable(),
+	details: z.string(),
-	testStrategy: z.string().nullable(),
+	testStrategy: z.string(),
-	priority: z.enum(['high', 'medium', 'low']).nullable(),
+	priority: z.enum(['high', 'medium', 'low']),
-	dependencies: z.array(z.number().int().positive()).nullable(),
+	dependencies: z.array(z.number()),
-	status: z.string().nullable()
+	status: z.string()
 });
 
 // Define the Zod schema for the ENTIRE expected AI response object
@@ -257,10 +257,15 @@ async function parsePRD(prdPath, tasksPath, numTasks, options = {}) {
 			return {
 				...task,
 				id: newId,
-				status: 'pending',
+				status: task.status || 'pending',
 				priority: task.priority || 'medium',
 				dependencies: Array.isArray(task.dependencies) ? task.dependencies : [],
-				subtasks: []
+				subtasks: [],
+				// Ensure all required fields have values (even if empty strings)
+				title: task.title || '',
+				description: task.description || '',
+				details: task.details || '',
+				testStrategy: task.testStrategy || ''
 			};
 		});
 

scripts/modules/task-manager/scope-adjustment.js

@@ -337,7 +337,7 @@ ${
 }
 
 Return a JSON object with a "subtasks" array. Each subtask should have:
-- id: Sequential number starting from 1
+- id: Sequential NUMBER starting from 1 (e.g., 1, 2, 3 - NOT "1", "2", "3")
 - title: Clear, specific title
 - description: Detailed description
 - dependencies: Array of dependency IDs as STRINGS (use format ["${task.id}.1", "${task.id}.2"] for siblings, or empty array [] for no dependencies)
@@ -345,7 +345,9 @@ Return a JSON object with a "subtasks" array. Each subtask should have:
 - status: "pending"
 - testStrategy: Testing approach
 
-IMPORTANT: Dependencies must be strings, not numbers!
+IMPORTANT: 
+- The 'id' field must be a NUMBER, not a string!
+- Dependencies must be strings, not numbers!
 
 Ensure the JSON is valid and properly formatted.`;
 
@@ -358,14 +360,14 @@ Ensure the JSON is valid and properly formatted.`;
 					description: z.string().min(10),
 					dependencies: z.array(z.string()),
 					details: z.string().min(20),
-					status: z.string().default('pending'),
+					status: z.string(),
-					testStrategy: z.string().nullable().default('')
+					testStrategy: z.string()
 				})
 			)
 		});
 
 		const aiResult = await generateObjectService({
-			role: 'main',
+			role: context.research ? 'research' : 'main',
 			session: context.session,
 			systemPrompt,
 			prompt,
@@ -377,14 +379,21 @@ Ensure the JSON is valid and properly formatted.`;
 
 		const generatedSubtasks = aiResult.mainResult.subtasks || [];
 
+		// Post-process generated subtasks to ensure defaults
+		const processedGeneratedSubtasks = generatedSubtasks.map((subtask) => ({
+			...subtask,
+			status: subtask.status || 'pending',
+			testStrategy: subtask.testStrategy || ''
+		}));
+
 		// Update task with preserved subtasks + newly generated ones
-		task.subtasks = [...preservedSubtasks, ...generatedSubtasks];
+		task.subtasks = [...preservedSubtasks, ...processedGeneratedSubtasks];
 
 		return {
 			updatedTask: task,
 			regenerated: true,
 			preserved: preservedSubtasks.length,
-			generated: generatedSubtasks.length
+			generated: processedGeneratedSubtasks.length
 		};
 	} catch (error) {
 		log(
@@ -457,6 +466,7 @@ ADJUSTMENT REQUIREMENTS:
 - description: Updated task description  
 - details: Updated implementation details
 - testStrategy: Updated test strategy
+- priority: Task priority ('low', 'medium', or 'high')
 
 Ensure the JSON is valid and properly formatted.`;
 
@@ -501,14 +511,11 @@ async function adjustTaskComplexity(
 			.string()
 			.min(1)
 			.describe('Updated testing approach for the adjusted scope'),
-		priority: z
+		priority: z.enum(['low', 'medium', 'high']).describe('Task priority level')
-			.enum(['low', 'medium', 'high'])
-			.optional()
-			.describe('Task priority level')
 	});
 
 	const aiResult = await generateObjectService({
-		role: 'main',
+		role: context.research ? 'research' : 'main',
 		session: context.session,
 		systemPrompt,
 		prompt,
@@ -520,10 +527,16 @@ async function adjustTaskComplexity(
 
 	const updatedTaskData = aiResult.mainResult;
 
+	// Ensure priority has a value (in case AI didn't provide one)
+	const processedTaskData = {
+		...updatedTaskData,
+		priority: updatedTaskData.priority || task.priority || 'medium'
+	};
+
 	return {
 		updatedTask: {
 			...task,
-			...updatedTaskData
+			...processedTaskData
 		},
 		telemetryData: aiResult.telemetryData
 	};

src/ai-providers/base-provider.js

@@ -1,4 +1,12 @@
-import { generateObject, generateText, streamText } from 'ai';
+import {
+	generateObject,
+	generateText,
+	streamText,
+	zodSchema,
+	JSONParseError,
+	NoObjectGeneratedError
+} from 'ai';
+import { jsonrepair } from 'jsonrepair';
 import { log } from '../../scripts/modules/utils.js';
 
 /**
@@ -206,8 +214,8 @@ export class BaseAIProvider {
 			const result = await generateObject({
 				model: client(params.modelId),
 				messages: params.messages,
-				schema: params.schema,
+				schema: zodSchema(params.schema),
-				mode: 'auto',
+				mode: params.mode || 'auto',
 				maxTokens: params.maxTokens,
 				temperature: params.temperature
 			});
@@ -226,6 +234,43 @@ export class BaseAIProvider {
 				}
 			};
 		} catch (error) {
+			// Check if this is a JSON parsing error that we can potentially fix
+			if (
+				NoObjectGeneratedError.isInstance(error) &&
+				JSONParseError.isInstance(error.cause) &&
+				error.cause.text
+			) {
+				log(
+					'warn',
+					`${this.name} generated malformed JSON, attempting to repair...`
+				);
+
+				try {
+					// Use jsonrepair to fix the malformed JSON
+					const repairedJson = jsonrepair(error.cause.text);
+					const parsed = JSON.parse(repairedJson);
+
+					log('info', `Successfully repaired ${this.name} JSON output`);
+
+					// Return in the expected format
+					return {
+						object: parsed,
+						usage: {
+							// Extract usage information from the error if available
+							inputTokens: error.usage?.promptTokens || 0,
+							outputTokens: error.usage?.completionTokens || 0,
+							totalTokens: error.usage?.totalTokens || 0
+						}
+					};
+				} catch (repairError) {
+					log(
+						'error',
+						`Failed to repair ${this.name} JSON: ${repairError.message}`
+					);
+					// Fall through to handleError with original error
+				}
+			}
+
 			this.handleError('object generation', error);
 		}
 	}

src/ai-providers/perplexity.js

@@ -44,4 +44,21 @@ export class PerplexityAIProvider extends BaseAIProvider {
 			this.handleError('client initialization', error);
 		}
 	}
+
+	/**
+	 * Override generateObject to use JSON mode for Perplexity
+	 *
+	 * NOTE: Perplexity models (especially sonar models) have known issues
+	 * generating valid JSON, particularly with array fields. They often
+	 * generate malformed JSON like "dependencies": , instead of "dependencies": []
+	 *
+	 * The base provider now handles JSON repair automatically for all providers.
+	 */
+	async generateObject(params) {
+		// Force JSON mode for Perplexity as it may help with reliability
+		return super.generateObject({
+			...params,
+			mode: 'json'
+		});
+	}
 }