chore: fix logging error in mcp stdio

.changeset/open-tips-notice.md

@@ -0,0 +1,5 @@
+---
+"task-master-ai": minor
+---
+
+Add 4.5 haiku and sonnet to supported models for claude-code and anthropic ai providers

apps/docs/capabilities/cli-root-commands.mdx

@@ -3,22 +3,6 @@ title: CLI Commands
 sidebarTitle: "CLI Commands"
 ---
 
-## Debug Mode
-
-For troubleshooting issues, you can enable debug mode to get detailed error information:
-
-```bash
-# Enable debug mode for any command
-DEBUG=true task-master <command>
-
-# Example with next command
-DEBUG=true task-master next
-```
-
-Debug mode shows full error messages, stack traces, and additional context. See the [Troubleshooting guide](/getting-started/troubleshooting) for more details.
-
-## Commands
-
 
 <AccordionGroup>
   <Accordion title="Parse PRD">

apps/docs/docs.json

@@ -33,7 +33,6 @@
 								]
 							},
 							"getting-started/api-keys",
-							"getting-started/troubleshooting",
 							"getting-started/faq",
 							"getting-started/contribute"
 						]

apps/docs/getting-started/faq.mdx

@@ -3,32 +3,7 @@ title: FAQ
 sidebarTitle: "FAQ"
 ---
 
-## Common Questions
+Coming soon. 
-
-### How do I get more detailed error information?
-
-Enable debug mode by setting the `DEBUG` environment variable:
-
-```bash
-DEBUG=true task-master <command>
-```
-
-This shows full error messages, stack traces, and additional context. See our [Troubleshooting guide](/getting-started/troubleshooting) for complete details.
-
-### Why are my commands failing silently?
-
-Task Master uses sanitized error messages by default to provide a clean user experience. Enable debug mode to see technical details about what went wrong.
-
-### How do I report a bug?
-
-When reporting issues, please:
-1. Enable debug mode and include the full error output
-2. Provide your environment details (OS, Node.js version, Task Master version)
-3. Include the command that failed and your project configuration
-
-See our [Troubleshooting guide](/getting-started/troubleshooting) for complete bug reporting guidelines.
-
-More frequently asked questions coming soon. 
 
 ## 💬 Getting Help
 

apps/docs/getting-started/troubleshooting.mdx

@@ -1,155 +0,0 @@
----
-title: Troubleshooting
-sidebarTitle: "Troubleshooting"
----
-
-This guide helps you troubleshoot common issues and get more detailed error information when using Task Master.
-
-## Debug Mode
-
-Task Master includes a debug mode that provides detailed error information including stack traces. This is especially useful when reporting issues or debugging problems.
-
-### Enabling Debug Mode
-
-Set the `DEBUG` environment variable to enable debug mode:
-
-```bash
-# Enable debug mode for a single command
-DEBUG=true task-master next
-
-# Or use '1' instead of 'true'
-DEBUG=1 task-master list
-
-# Enable for entire session (bash/zsh)
-export DEBUG=true
-task-master next  # Now shows detailed errors
-
-# Enable for entire session (fish shell)
-set -x DEBUG true
-task-master next
-```
-
-### What Debug Mode Shows
-
-With debug mode enabled, errors will include:
-
-- **Full error messages**: Complete technical details instead of sanitized user messages
-- **Stack traces**: Detailed information about where errors occurred
-- **Error context**: Additional metadata about the failed operation
-- **Error chains**: If an error was caused by another error, both are shown
-
-### Example Output
-
-**Normal mode (default):**
-```
-Error: Unable to read tasks file
-```
-
-**Debug mode:**
-```
-TaskMasterError[FILE_READ_ERROR]: Unable to read tasks file
-
-Stack trace:
-  at TaskService.loadTasks (/path/to/task-service.ts:123)
-  at NextCommand.execute (/path/to/next.command.ts:45)
-  ...
-```
-
-## Common Issues
-
-### File Permission Errors
-
-**Problem**: Commands fail with file access errors
-
-**Solution**: 
-1. Check file permissions in `.taskmaster/` directory
-2. Ensure your user has write access to the project directory
-3. Run with debug mode to see the exact file path causing issues
-
-### API Connection Issues
-
-**Problem**: Commands timeout or fail when using API storage
-
-**Solutions**:
-1. Check your internet connection
-2. Verify API keys are correctly configured
-3. Check if your organization has firewall restrictions
-4. Use debug mode to see specific network errors
-
-### Task File Corruption
-
-**Problem**: Commands report invalid task data
-
-**Solutions**:
-1. Check if `.taskmaster/tasks/tasks.json` is valid JSON
-2. Restore from backup if available
-3. Run `task-master generate` to regenerate task files from main data
-
-### Model Configuration Issues
-
-**Problem**: AI-powered commands fail
-
-**Solutions**:
-1. Verify API keys are set: `task-master models`
-2. Check if the configured model is available
-3. Try switching to a different model
-4. Use debug mode to see specific API error messages
-
-## Getting Help
-
-When reporting issues, please:
-
-1. **Enable debug mode** and include the full error output
-2. **Specify your environment**:
-   - Operating system and version
-   - Node.js version (`node --version`)
-   - Task Master version (`task-master --version`)
-3. **Include relevant context**:
-   - Command that failed
-   - Project configuration
-   - Recent changes to your setup
-
-### Support Channels
-
-- **GitHub Issues**: [Report bugs and request features](https://github.com/eyaltoledano/claude-task-master/issues)
-- **Discord**: [Join our community for help](https://discord.gg/taskmasterai)
-- **GitHub Discussions**: [Ask questions and share ideas](https://github.com/eyaltoledano/claude-task-master/discussions)
-
-## Advanced Debugging
-
-### Inspecting Task Data
-
-If you suspect task data corruption:
-
-```bash
-# View raw task data
-cat .taskmaster/tasks/tasks.json | jq '.'
-
-# Check specific task
-task-master show <task-id> --format=json
-```
-
-### Configuration Debugging
-
-Check your current configuration:
-
-```bash
-# View model configuration
-task-master models
-
-# Check if initialization is complete
-ls -la .taskmaster/
-```
-
-### Network Debugging
-
-For API storage issues:
-
-```bash
-# Test with curl (replace with your API endpoint)
-curl -H "Authorization: Bearer $ANTHROPIC_API_KEY" \
-     -H "Content-Type: application/json" \
-     https://api.anthropic.com/v1/models
-```
-
-Remember to enable debug mode (`DEBUG=true`) when troubleshooting to get the most helpful error information.

docs/models.md

@@ -1,4 +1,4 @@
-# Available Models as of October 5, 2025
+# Available Models as of October 18, 2025
 
 ## Main Models
 
@@ -8,8 +8,11 @@
 | anthropic   | claude-opus-4-20250514                         | 0.725     | 15         | 75          |
 | anthropic   | claude-3-7-sonnet-20250219                     | 0.623     | 3          | 15          |
 | anthropic   | claude-3-5-sonnet-20241022                     | 0.49      | 3          | 15          |
+| anthropic   | claude-sonnet-4-5-20250929                     | 0.73      | 3          | 15          |
+| anthropic   | claude-haiku-4-5-20251001                      | 0.45      | 1          | 5           |
 | claude-code | opus                                           | 0.725     | 0          | 0           |
 | claude-code | sonnet                                         | 0.727     | 0          | 0           |
+| claude-code | haiku                                          | 0.45      | 0          | 0           |
 | codex-cli   | gpt-5                                          | 0.749     | 0          | 0           |
 | codex-cli   | gpt-5-codex                                    | 0.749     | 0          | 0           |
 | mcp         | mcp-sampling                                   | —         | 0          | 0           |
@@ -102,6 +105,7 @@
 | ----------- | -------------------------------------------- | --------- | ---------- | ----------- |
 | claude-code | opus                                         | 0.725     | 0          | 0           |
 | claude-code | sonnet                                       | 0.727     | 0          | 0           |
+| claude-code | haiku                                        | 0.45      | 0          | 0           |
 | codex-cli   | gpt-5                                        | 0.749     | 0          | 0           |
 | codex-cli   | gpt-5-codex                                  | 0.749     | 0          | 0           |
 | mcp         | mcp-sampling                                 | —         | 0          | 0           |
@@ -142,8 +146,11 @@
 | anthropic   | claude-opus-4-20250514                         | 0.725     | 15         | 75          |
 | anthropic   | claude-3-7-sonnet-20250219                     | 0.623     | 3          | 15          |
 | anthropic   | claude-3-5-sonnet-20241022                     | 0.49      | 3          | 15          |
+| anthropic   | claude-sonnet-4-5-20250929                     | 0.73      | 3          | 15          |
+| anthropic   | claude-haiku-4-5-20251001                      | 0.45      | 1          | 5           |
 | claude-code | opus                                           | 0.725     | 0          | 0           |
 | claude-code | sonnet                                         | 0.727     | 0          | 0           |
+| claude-code | haiku                                          | 0.45      | 0          | 0           |
 | codex-cli   | gpt-5                                          | 0.749     | 0          | 0           |
 | codex-cli   | gpt-5-codex                                    | 0.749     | 0          | 0           |
 | mcp         | mcp-sampling                                   | —         | 0          | 0           |

mcp-server/src/logger.js

@@ -78,7 +78,7 @@ function log(level, ...args) {
 		// is responsible for directing logs correctly (e.g., to stderr)
 		// during tool execution without upsetting the client connection.
 		// Logs outside of tool execution (like startup) will go to stdout.
-		console.log(prefix, ...coloredArgs);
+		console.error(prefix, ...coloredArgs);
 	}
 }
 

packages/tm-core/src/logger/logger.ts

@@ -25,7 +25,7 @@ export interface LoggerConfig {
 export class Logger {
 	private config: Required<LoggerConfig>;
 	private static readonly DEFAULT_CONFIG: Required<LoggerConfig> = {
-		level: LogLevel.WARN,
+		level: LogLevel.SILENT,
 		silent: false,
 		prefix: '',
 		timestamp: false,

scripts/modules/config-manager.js

@@ -307,6 +307,20 @@ function validateProviderModelCombination(providerName, modelId) {
 	);
 }
 
+/**
+ * Gets the list of supported model IDs for a given provider from supported-models.json
+ * @param {string} providerName - The name of the provider (e.g., 'claude-code', 'anthropic')
+ * @returns {string[]} Array of supported model IDs, or empty array if provider not found
+ */
+export function getSupportedModelsForProvider(providerName) {
+	if (!MODEL_MAP[providerName]) {
+		return [];
+	}
+	return MODEL_MAP[providerName]
+		.filter((model) => model.supported !== false)
+		.map((model) => model.id);
+}
+
 /**
  * Validates Claude Code AI provider custom settings
  * @param {object} settings The settings to validate

scripts/modules/supported-models.json

@@ -43,6 +43,28 @@
 			"allowed_roles": ["main", "fallback"],
 			"max_tokens": 8192,
 			"supported": true
+		},
+		{
+			"id": "claude-sonnet-4-5-20250929",
+			"swe_score": 0.73,
+			"cost_per_1m_tokens": {
+				"input": 3.0,
+				"output": 15.0
+			},
+			"allowed_roles": ["main", "fallback"],
+			"max_tokens": 64000,
+			"supported": true
+		},
+		{
+			"id": "claude-haiku-4-5-20251001",
+			"swe_score": 0.45,
+			"cost_per_1m_tokens": {
+				"input": 1.0,
+				"output": 5.0
+			},
+			"allowed_roles": ["main", "fallback"],
+			"max_tokens": 200000,
+			"supported": true
 		}
 	],
 	"claude-code": [
@@ -67,6 +89,17 @@
 			"allowed_roles": ["main", "fallback", "research"],
 			"max_tokens": 64000,
 			"supported": true
+		},
+		{
+			"id": "haiku",
+			"swe_score": 0.45,
+			"cost_per_1m_tokens": {
+				"input": 0,
+				"output": 0
+			},
+			"allowed_roles": ["main", "fallback", "research"],
+			"max_tokens": 200000,
+			"supported": true
 		}
 	],
 	"codex-cli": [

src/ai-providers/claude-code.js

@@ -12,7 +12,10 @@
 
 import { createClaudeCode } from 'ai-sdk-provider-claude-code';
 import { BaseAIProvider } from './base-provider.js';
-import { getClaudeCodeSettingsForCommand } from '../../scripts/modules/config-manager.js';
+import {
+	getClaudeCodeSettingsForCommand,
+	getSupportedModelsForProvider
+} from '../../scripts/modules/config-manager.js';
 import { execSync } from 'child_process';
 import { log } from '../../scripts/modules/utils.js';
 
@@ -24,14 +27,24 @@ let _claudeCliAvailable = null;
  *
  * Features:
  * - No API key required (uses local Claude Code CLI)
- * - Supports 'sonnet' and 'opus' models
+ * - Supported models loaded from supported-models.json
  * - Command-specific configuration support
  */
 export class ClaudeCodeProvider extends BaseAIProvider {
 	constructor() {
 		super();
 		this.name = 'Claude Code';
-		this.supportedModels = ['sonnet', 'opus'];
+		// Load supported models from supported-models.json
+		this.supportedModels = getSupportedModelsForProvider('claude-code');
+
+		// Validate that models were loaded successfully
+		if (this.supportedModels.length === 0) {
+			log(
+				'warn',
+				'No supported models found for claude-code provider. Check supported-models.json configuration.'
+			);
+		}
+
 		// Claude Code requires explicit JSON schema mode
 		this.needsExplicitJsonSchema = true;
 		// Claude Code does not support temperature parameter

src/ai-providers/codex-cli.js

@@ -10,7 +10,10 @@ import { createCodexCli } from 'ai-sdk-provider-codex-cli';
 import { BaseAIProvider } from './base-provider.js';
 import { execSync } from 'child_process';
 import { log } from '../../scripts/modules/utils.js';
-import { getCodexCliSettingsForCommand } from '../../scripts/modules/config-manager.js';
+import {
+	getCodexCliSettingsForCommand,
+	getSupportedModelsForProvider
+} from '../../scripts/modules/config-manager.js';
 
 export class CodexCliProvider extends BaseAIProvider {
 	constructor() {
@@ -20,8 +23,17 @@ export class CodexCliProvider extends BaseAIProvider {
 		this.needsExplicitJsonSchema = false;
 		// Codex CLI does not support temperature parameter
 		this.supportsTemperature = false;
-		// Restrict to supported models for OAuth subscription usage
+		// Load supported models from supported-models.json
-		this.supportedModels = ['gpt-5', 'gpt-5-codex'];
+		this.supportedModels = getSupportedModelsForProvider('codex-cli');
+
+		// Validate that models were loaded successfully
+		if (this.supportedModels.length === 0) {
+			log(
+				'warn',
+				'No supported models found for codex-cli provider. Check supported-models.json configuration.'
+			);
+		}
+
 		// CLI availability check cache
 		this._codexCliChecked = false;
 		this._codexCliAvailable = null;

tests/integration/claude-code-error-handling.test.js

@@ -43,9 +43,9 @@ describe('Claude Code Error Handling', () => {
 
 		// These should work even if CLI is not available
 		expect(provider.name).toBe('Claude Code');
-		expect(provider.getSupportedModels()).toEqual(['sonnet', 'opus']);
+		expect(provider.getSupportedModels()).toEqual(['opus', 'sonnet', 'haiku']);
 		expect(provider.isModelSupported('sonnet')).toBe(true);
-		expect(provider.isModelSupported('haiku')).toBe(false);
+		expect(provider.isModelSupported('haiku')).toBe(true);
 		expect(provider.isRequiredApiKey()).toBe(false);
 		expect(() => provider.validateAuth()).not.toThrow();
 	});

tests/integration/claude-code-optional.test.js

@@ -40,14 +40,14 @@ describe('Claude Code Integration (Optional)', () => {
 	it('should create a working provider instance', () => {
 		const provider = new ClaudeCodeProvider();
 		expect(provider.name).toBe('Claude Code');
-		expect(provider.getSupportedModels()).toEqual(['sonnet', 'opus']);
+		expect(provider.getSupportedModels()).toEqual(['opus', 'sonnet', 'haiku']);
 	});
 
 	it('should support model validation', () => {
 		const provider = new ClaudeCodeProvider();
 		expect(provider.isModelSupported('sonnet')).toBe(true);
 		expect(provider.isModelSupported('opus')).toBe(true);
-		expect(provider.isModelSupported('haiku')).toBe(false);
+		expect(provider.isModelSupported('haiku')).toBe(true);
 		expect(provider.isModelSupported('unknown')).toBe(false);
 	});
 

tests/unit/ai-providers/claude-code.test.js

@@ -28,6 +28,14 @@ jest.unstable_mockModule('../../../src/ai-providers/base-provider.js', () => ({
 	}
 }));
 
+// Mock config getters
+jest.unstable_mockModule('../../../scripts/modules/config-manager.js', () => ({
+	getClaudeCodeSettingsForCommand: jest.fn(() => ({})),
+	getSupportedModelsForProvider: jest.fn(() => ['opus', 'sonnet', 'haiku']),
+	getDebugFlag: jest.fn(() => false),
+	getLogLevel: jest.fn(() => 'info')
+}));
+
 // Import after mocking
 const { ClaudeCodeProvider } = await import(
 	'../../../src/ai-providers/claude-code.js'
@@ -96,13 +104,13 @@ describe('ClaudeCodeProvider', () => {
 	describe('model support', () => {
 		it('should return supported models', () => {
 			const models = provider.getSupportedModels();
-			expect(models).toEqual(['sonnet', 'opus']);
+			expect(models).toEqual(['opus', 'sonnet', 'haiku']);
 		});
 
 		it('should check if model is supported', () => {
 			expect(provider.isModelSupported('sonnet')).toBe(true);
 			expect(provider.isModelSupported('opus')).toBe(true);
-			expect(provider.isModelSupported('haiku')).toBe(false);
+			expect(provider.isModelSupported('haiku')).toBe(true);
 			expect(provider.isModelSupported('unknown')).toBe(false);
 		});
 	});

tests/unit/ai-providers/codex-cli.test.js

@@ -20,6 +20,7 @@ jest.unstable_mockModule('ai-sdk-provider-codex-cli', () => ({
 // Mock config getters
 jest.unstable_mockModule('../../../scripts/modules/config-manager.js', () => ({
 	getCodexCliSettingsForCommand: jest.fn(() => ({ allowNpx: true })),
+	getSupportedModelsForProvider: jest.fn(() => ['gpt-5', 'gpt-5-codex']),
 	// Provide commonly imported getters to satisfy other module imports if any
 	getDebugFlag: jest.fn(() => false),
 	getLogLevel: jest.fn(() => 'info')