- Unified Service: Introduced 'scripts/modules/ai-services-unified.js' to centralize AI interactions using provider modules ('src/ai-providers/') and the Vercel AI SDK.
- Provider Modules: Implemented 'anthropic.js' and 'perplexity.js' wrappers for Vercel SDK.
- 'updateSubtaskById' Fix: Refactored the AI call within 'updateSubtaskById' to use 'generateTextService' from the unified layer, resolving runtime errors related to parameter passing and streaming. This serves as the pattern for refactoring other AI calls in 'scripts/modules/task-manager/'.
- Task Status: Marked Subtask 61.19 as 'done'.
- Rules: Added new 'ai-services.mdc' rule.
This centralizes AI logic, replacing previous direct SDK calls and custom implementations. API keys are resolved via 'resolveEnvVariable' within the service layer. The refactoring of 'updateSubtaskById' establishes the standard approach for migrating other AI-dependent functions in the task manager module to use the unified service.
Relates to Task 61.
BREAKING CHANGE: Taskmaster now requires a `.taskmasterconfig` file for model/parameter settings. Environment variables (except API keys) are no longer used for overrides.
- Throws an error if `.taskmasterconfig` is missing, guiding user to run `task-master models --setup`." -m "- Removed env var checks from config getters in `config-manager.js`." -m "- Updated `env.example` to remove obsolete variables." -m "- Refined missing config file error message in `commands.js`.
The interactive model setup triggered by `task-master models --setup` was previously attempting to call non-existent setter functions (`setMainModel`, etc.) in `config-manager.js`, leading to errors and preventing configuration updates.
This commit refactors the `--setup` logic within the `models` command handler in `scripts/modules/commands.js`. It now correctly:
- Loads the current configuration using `getConfig()`." -m "- Updates the appropriate sections of the loaded configuration object based on user selections from `inquirer`." -m "- Saves the modified configuration using the existing `writeConfig()` function from `config-manager.js`." -m "- Handles disabling the fallback model correctly."
This commit centralizes configuration and environment variable access across various modules by consistently utilizing getters from scripts/modules/config-manager.js. This replaces direct access to process.env and the global CONFIG object, leading to improved consistency, maintainability, testability, and better handling of session-specific configurations within the MCP context.
Key changes include:
- Centralized Getters: Replaced numerous instances of process.env.* and CONFIG.* with corresponding getter functions (e.g., getLogLevel, getMainModelId, getResearchMaxTokens, getMainTemperature, isApiKeySet, getDebugFlag, getDefaultSubtasks).
- Session Awareness: Ensured that the session object is passed to config getters where necessary, particularly within AI service calls (ai-services.js, add-task.js) and error handling (ai-services.js), allowing for session-specific environment overrides.
- API Key Checks: Standardized API key availability checks using isApiKeySet() instead of directly checking process.env.* (e.g., for Perplexity in commands.js and ai-services.js).
- Client Instantiation Cleanup: Removed now-redundant/obsolete local client instantiation functions (getAnthropicClient, getPerplexityClient) from ai-services.js and the global Anthropic client initialization from dependency-manager.js. Client creation should now rely on the config manager and factory patterns.
- Consistent Debug Flag Usage: Standardized calls to getDebugFlag() in commands.js, removing potentially unnecessary null arguments.
- Accurate Progress Calculation: Updated AI stream progress reporting (ai-services.js, add-task.js) to use getMainMaxTokens(session) for more accurate calculations.
- Minor Cleanup: Removed unused import from scripts/modules/commands.js.
Specific module updates:
- :
- Uses getLogLevel() instead of process.env.LOG_LEVEL.
- :
- Replaced direct env/config access for model IDs, tokens, temperature, API keys, and default subtasks with appropriate getters.
- Passed session to handleClaudeError.
- Removed local getPerplexityClient and getAnthropicClient functions.
- Updated progress calculations to use getMainMaxTokens(session).
- :
- Uses isApiKeySet('perplexity') for API key checks.
- Uses getDebugFlag() consistently for debug checks.
- Removed unused import.
- :
- Removed global Anthropic client initialization.
- :
- Uses config getters (getResearch..., getMain...) for Perplexity and Claude API call parameters, preserving customEnv override logic.
This refactoring also resolves a potential SyntaxError: Identifier 'getPerplexityClient' has already been declared by removing the duplicated/obsolete function definition previously present in ai-services.js.
This commit focuses on standardizing configuration and API key access patterns across key modules as part of subtask 61.34.
Key changes include:
- Refactored `ai-services.js` to remove global AI clients and use `resolveEnvVariable` for API key checks. Client instantiation now relies on `getAnthropicClient`/`getPerplexityClient` accepting a session object.
- Refactored `task-manager.js` (`analyzeTaskComplexity` function) to use the unified `generateTextService` from `ai-services-unified.js`, removing direct AI client calls.
- Replaced direct `process.env` access for model parameters and other configurations (`PERPLEXITY_MODEL`, `CONFIG.*`) in `task-manager.js` with calls to the appropriate getters from `config-manager.js` (e.g., `getResearchModelId(session)`, `getMainMaxTokens(session)`).
- Ensured `utils.js` (`resolveEnvVariable`) correctly handles potentially undefined session objects.
- Updated function signatures where necessary to propagate the `session` object for correct context-aware configuration/key retrieval.
This moves towards the goal of using `ai-client-factory.js` and `ai-services-unified.js` as the standard pattern for AI interactions and centralizing configuration management through `config-manager.js`.
- Integrate for Grok models and for OpenRouter into the AI client factory ().
- Install necessary provider dependencies (, , and other related packages, updated core).
- Update environment variable checks () and client creation logic () for the new providers.
- Add and correct unit tests in to cover xAI and OpenRouter instantiation, error handling, and environment variable resolution.
- Corrected mock paths and names in tests to align with official package names.
- Verify all tests (28 total) pass for .
- Confirm test coverage remains high (~90%) after additions.
* prompt engineering prd breakdown
* chore: add back important elements of the parsePRD prompt
---------
Co-authored-by: chen kinnrot <chen.kinnrot@lemonade.com>
Introduces a configurable fallback model and adds support for additional AI provider API keys in the environment setup.
- **Add Fallback Model Configuration (.taskmasterconfig):**
- Implemented a new section in .
- Configured as the default fallback model, enhancing resilience if the primary model fails.
- **Update Default Model Configuration (.taskmasterconfig):**
- Changed the default model to .
- Changed the default model to .
- **Add API Key Examples (assets/env.example):**
- Added example environment variables for:
- (for OpenAI/OpenRouter)
- (for Google Gemini)
- (for XAI Grok)
- Included format comments for clarity.
Refactored `config-manager.js` to handle different execution contexts (CLI vs. MCP) and fixed related Jest tests.
- Modified `readConfig` and `writeConfig` to accept an optional `explicitRoot` parameter, allowing explicit path specification (e.g., from MCP) while retaining automatic project root finding for CLI usage.
- Updated getter/setter functions (`getMainProvider`, `setMainModel`, etc.) to accept and propagate the `explicitRoot`.
- Resolved Jest testing issues for dynamic imports by using `jest.unstable_mockModule` for `fs` and `chalk` dependencies *before* the dynamic `import()`.
- Corrected console error assertions in tests to match exact logged messages.
- Updated `.cursor/rules/tests.mdc` with guidelines for `jest.unstable_mockModule` and precise console assertions.
Addresses `process.stdout.clearLine is not a function` error when running AI-dependent commands non-interactively (e.g., `update-subtask`).
Adds `process.stdout.isTTY` check before attempting to use terminal-specific output manipulations.
feat: Implement initial config manager for AI models
Adds `scripts/modules/config-manager.js` to handle reading/writing model selections from/to `.taskmasterconfig`.
Implements core functions: findProjectRoot, read/writeConfig, validateModel, get/setModel.
Defines valid model lists. Completes initial work for Subtask 61.1.
The function used terminal manipulation functions
(like , ) for the CLI
streaming progress indicator. This caused errors when Task Master commands
involving AI streaming were run in non-interactive terminals (e.g., via
output redirection, some CI environments, or integrated terminals).
This commit adds a check for to the condition
that controls the display of the CLI progress indicator, ensuring these
functions are only called when standard output is a fully interactive TTY.
Improves the quality and relevance of research-backed AI operations:
- Tweaks Perplexity AI calls to use max input tokens (8700), temperature 0.1, high context size, and day-fresh search recency.
- Adds a system prompt to guide Perplexity research output.
Docs:
- Updates CLI examples in taskmaster.mdc to use ANSI-C quoting ($'...') for multi-line prompts, ensuring they work correctly in bash/zsh.
Centralizes init command logic within the main CLI structure. The action handler in commands.js now directly calls initializeProject from the init.js module, resolving issues with argument parsing (like -y) and removing the need for the separate bin/task-master-init.js executable. Updates package.json and bin/task-master.js accordingly.
- Add support for --title/-t and --description/-d flags in add-task command
- Fix validation for manual creation mode (title + description)
- Implement proper testing for both prompt and manual creation modes
- Update testing documentation with Commander.js testing best practices
- Add guidance on handling variable hoisting and module initialization issues
Changeset: brave-doors-open.md
- Fix expand-all command bugs that caused NaN errors with --all option and JSON formatting errors with research enabled
- Improve error handling to provide clear feedback when subtask generation fails
- Include task IDs and actionable suggestions in error messages
- Refactor for more reliable project root detection, particularly when running within integrated environments like Cursor IDE. Includes deriving root from script path and avoiding fallback to '/'.
- Enhance error handling in :
- Add detailed debug information (paths searched, CWD, etc.) to the error message when is not found in the provided project root.
- Improve clarity of error messages and potential solutions.
- Add verbose logging in to trace session object content and the finally resolved project root path, aiding in debugging path-related issues.
- Add default values for and to the example environment configuration.
- Add cancelled status to UI module for marking tasks cancelled without deletion
- Improve MCP server resource documentation with implementation examples
- Update architecture.mdc with detailed resource management info
- Add comprehensive resource handling guide to mcp.mdc
- Update changeset to reflect new features and documentation
- Mark task 23.6 as cancelled (MCP SDK integration no longer needed)
- Complete task 23.12 (structured logging system)
Slight refactor moving numTasks and outputPath to top with the other variables. Eliminates duplication, and keeps us from having to check path twice.
Resolves#65
Implemented LRU caching for the function to improve performance for repeated requests.
Key changes include:
- Added dependency.
- Introduced a reusable utility function in leveraging a .
- Refactored in to use the caching utility with a key based on task path, filter, and subtask flag.
- Modified to include the boolean flag in the final JSON response structure, nesting the original data under a key.
- Added function and corresponding MCP tool () for monitoring cache performance.
- Improved error handling in for cases where is not found.
This addresses the previous issue of the empty task list likely caused by stale cache entries and provides clear visibility into whether a response is served from the cache.
Relates to #23.9
This commit introduces a major refactoring of the MCP server implementation to prioritize direct function calls over CLI execution, enhancing performance and reliability. It also includes substantial updates to documentation for consistency and interlinking.
**MCP Server & Core Logic Refactoring:**
1. **Introduce Direct Function Wrappers ():**
* Created to house direct wrappers for core Task Master functions (imported from ).
* Implemented as the first wrapper, calling .
* Added utility within to centralize file location logic, removing duplication.
* Established the map for registering these wrappers.
2. **Enhance MCP Utilities ():**
* Added : A primary utility function to streamline MCP tool methods. It handles logging, argument processing (incl. project root normalization), calling the direct action function (e.g., ), processing results via , and formatting the final MCP response.
* Added : Standardizes processing of objects returned by direct function wrappers.
* Added : Filters sensitive/large fields (like , ) from responses sent to the MCP client.
* Added , , to support the new workflow.
* Refactored to use internally, simplifying its usage (though it's now primarily a fallback).
3. **Update MCP Tools (, ):**
* Refactored to use with , significantly simplifying the tool's method.
* Updated (initially) to use the improved (Note: further refactoring to use a direct wrapper for would follow the pattern).
**Documentation Enhancements:**
4. **Comprehensive Interlinking:** Added links across rule files (, , , , , ) to connect related guidelines, improving navigation.
5. **Standardize CLI Syntax ():** Removed legacy ℹ️ Initialized Perplexity client with OpenAI compatibility layer examples, reinforcing ℹ️ Initialized Perplexity client with OpenAI compatibility layer
_____ _ __ __ _
|_ _|_ _ ___| | __ | \/ | __ _ ___| |_ ___ _ __
| |/ _` / __| |/ / | |\/| |/ _` / __| __/ _ \ '__|
| | (_| \__ \ < | | | | (_| \__ \ || __/ |
|_|\__,_|___/_|\_\ |_| |_|\__,_|___/\__\___|_|
by https://x.com/eyaltoledano
╭────────────────────────────────────────────╮
│ │
│ Version: 0.9.30 Project: Task Master │
│ │
╰────────────────────────────────────────────╯
╭─────────────────────╮
│ │
│ Task Master CLI │
│ │
╰─────────────────────╯
╭───────────────────╮
│ Task Generation │
╰───────────────────╯
parse-prd --input=<file.txt> [--tasks=10] Generate tasks from a PRD document
generate Create individual task files from tasks…
╭───────────────────╮
│ Task Management │
╰───────────────────╯
list [--status=<status>] [--with-subtas… List all tasks with their status
set-status --id=<id> --status=<status> Update task status (done, pending, etc.)
update --from=<id> --prompt="<context>" Update tasks based on new requirements
add-task --prompt="<text>" [--dependencies=… Add a new task using AI
add-dependency --id=<id> --depends-on=<id> Add a dependency to a task
remove-dependency --id=<id> --depends-on=<id> Remove a dependency from a task
╭──────────────────────────╮
│ Task Analysis & Detail │
╰──────────────────────────╯
analyze-complexity [--research] [--threshold=5] Analyze tasks and generate expansion re…
complexity-report [--file=<path>] Display the complexity analysis report
expand --id=<id> [--num=5] [--research] [… Break down tasks into detailed subtasks
expand --all [--force] [--research] Expand all pending tasks with subtasks
clear-subtasks --id=<id> Remove subtasks from specified tasks
╭─────────────────────────────╮
│ Task Navigation & Viewing │
╰─────────────────────────────╯
next Show the next task to work on based on …
show <id> Display detailed information about a sp…
╭─────────────────────────╮
│ Dependency Management │
╰─────────────────────────╯
validate-dependenci… Identify invalid dependencies without f…
fix-dependencies Fix invalid dependencies automatically
╭─────────────────────────╮
│ Environment Variables │
╰─────────────────────────╯
ANTHROPIC_API_KEY Your Anthropic API key Required
MODEL Claude model to use Default: claude-3-7-sonn…
MAX_TOKENS Maximum tokens for responses Default: 4000
TEMPERATURE Temperature for model responses Default: 0.7
PERPLEXITY_API_KEY Perplexity API key for research Optional
PERPLEXITY_MODEL Perplexity model to use Default: sonar-pro
DEBUG Enable debug logging Default: false
LOG_LEVEL Console output level (debug,info,warn,error) Default: info
DEFAULT_SUBTASKS Default number of subtasks to generate Default: 3
DEFAULT_PRIORITY Default task priority Default: medium
PROJECT_NAME Project name displayed in UI Default: Task Master
_____ _ __ __ _
|_ _|_ _ ___| | __ | \/ | __ _ ___| |_ ___ _ __
| |/ _` / __| |/ / | |\/| |/ _` / __| __/ _ \ '__|
| | (_| \__ \ < | | | | (_| \__ \ || __/ |
|_|\__,_|___/_|\_\ |_| |_|\__,_|___/\__\___|_|
by https://x.com/eyaltoledano
╭────────────────────────────────────────────╮
│ │
│ Version: 0.9.30 Project: Task Master │
│ │
╰────────────────────────────────────────────╯
╭─────────────────────╮
│ │
│ Task Master CLI │
│ │
╰─────────────────────╯
╭───────────────────╮
│ Task Generation │
╰───────────────────╯
parse-prd --input=<file.txt> [--tasks=10] Generate tasks from a PRD document
generate Create individual task files from tasks…
╭───────────────────╮
│ Task Management │
╰───────────────────╯
list [--status=<status>] [--with-subtas… List all tasks with their status
set-status --id=<id> --status=<status> Update task status (done, pending, etc.)
update --from=<id> --prompt="<context>" Update tasks based on new requirements
add-task --prompt="<text>" [--dependencies=… Add a new task using AI
add-dependency --id=<id> --depends-on=<id> Add a dependency to a task
remove-dependency --id=<id> --depends-on=<id> Remove a dependency from a task
╭──────────────────────────╮
│ Task Analysis & Detail │
╰──────────────────────────╯
analyze-complexity [--research] [--threshold=5] Analyze tasks and generate expansion re…
complexity-report [--file=<path>] Display the complexity analysis report
expand --id=<id> [--num=5] [--research] [… Break down tasks into detailed subtasks
expand --all [--force] [--research] Expand all pending tasks with subtasks
clear-subtasks --id=<id> Remove subtasks from specified tasks
╭─────────────────────────────╮
│ Task Navigation & Viewing │
╰─────────────────────────────╯
next Show the next task to work on based on …
show <id> Display detailed information about a sp…
╭─────────────────────────╮
│ Dependency Management │
╰─────────────────────────╯
validate-dependenci… Identify invalid dependencies without f…
fix-dependencies Fix invalid dependencies automatically
╭─────────────────────────╮
│ Environment Variables │
╰─────────────────────────╯
ANTHROPIC_API_KEY Your Anthropic API key Required
MODEL Claude model to use Default: claude-3-7-sonn…
MAX_TOKENS Maximum tokens for responses Default: 4000
TEMPERATURE Temperature for model responses Default: 0.7
PERPLEXITY_API_KEY Perplexity API key for research Optional
PERPLEXITY_MODEL Perplexity model to use Default: sonar-pro
DEBUG Enable debug logging Default: false
LOG_LEVEL Console output level (debug,info,warn,error) Default: info
DEFAULT_SUBTASKS Default number of subtasks to generate Default: 3
DEFAULT_PRIORITY Default task priority Default: medium
PROJECT_NAME Project name displayed in UI Default: Task Master as the primary CLI.
6. **Add MCP Architecture & Workflow Sections:** Added detailed sections in and explaining MCP server structure and the workflow for adding new MCP tool integrations using the direct function pattern.
7. **Clarify MCP Role:** Updated and to better explain the MCP server's role and its specific utilities.
Overall, this establishes a cleaner, more performant, and maintainable pattern for integrating Task Master functionality with the MCP server, supported by improved documentation.
Ensures that the 'details' field, which can be updated via 'update-subtask', is correctly rendered when viewing a specific subtask.
fix(test): Remove empty describe block causing Jest error
Removes a redundant block in that contained a hook but no tests.
chore: Add npm script
This commit fixes an issue with the dependency management system where
dependencies between subtasks of the same parent task were not being
handled correctly. Previously, when trying to add a dependency between
subtasks (e.g., making 23.8 depend on 23.13), the system incorrectly
interpreted it as a circular dependency of the parent task depending
on itself.
Changes:
- Modified add-dependency and remove-dependency commands to preserve
string format for subtask IDs containing dots instead of converting
them to integers
- Updated dependency detection logic to properly handle string-based
subtask IDs
- Added specific tests verifying both successful dependency chains
and circular dependency detection between subtasks of the same parent
The fix ensures proper task ordering for development workflows where
related subtasks must be completed in sequence (like establishing tests
before implementing a feature). It maintains backward compatibility with
existing task structures while correctly enforcing dependency chains
within subtask groups.
Tests:
- Added tests for valid dependency chains between subtasks of the same parent
- Added tests for circular dependency detection in subtask relationships
- Added specific test for the 23.8->23.13->23.10 subtask dependency chain
Resolves the issue where the suggested task development workflow couldn't
be properly enforced through the dependency management system.
- Add support for comma-separated subtask IDs in remove-subtask command
- Implement MCP configuration in project initialization
- Add package update notification system with version comparison
- Improve command documentation with boolean flag conventions
- Add comprehensive error handling for unknown options
- Update help text with better examples and formatting
- Implement proper validation for command inputs
- Add global error handling patterns with helpful user messages
This commit fixes several issues with command line interface error handling:
1. Fix inconsistent behavior between --no-generate and --skip-generate:
- Standardized on --skip-generate across all commands
- Updated bin/task-master.js to use --skip-generate instead of --no-generate
- Modified add-subtask and remove-subtask commands to use --skip-generate
2. Enhance error handling for unknown options:
- Removed .allowUnknownOption() from commands to properly detect unknown options
- Added global error handler in bin/task-master.js for unknown commands/options
- Added command-specific error handlers with helpful error messages
3. Improve user experience with better help messages:
- Added helper functions to display formatted command help on errors
- Created command-specific help displays for add-subtask and remove-subtask
- Show available options when encountering unknown options
4. Update MCP server configuration:
- Modified .cursor/mcp.json to use node ./mcp-server/server.js directly
- Removed npx -y usage for more reliable execution
5. Other minor improvements:
- Adjusted column width for task ID display in UI
- Updated version number in package-lock.json to 0.9.30
This resolves issues where users would see confusing error messages like
'error: unknown option --generate' when using an incorrect flag."
Preserving completed subtasks exactly as they are
Building upon what has already been done
Creating new subtasks instead of modifying completed ones
Making new subtasks specific and targeted
Added specific instructions to the Perplexity AI system message to emphasize preserving completed subtasks
Added an informative boxed message to the user explaining how completed subtasks will be handled during the update process
Added emphatic instructions in the user prompts to both Claude and Perplexity to highlight completed subtasks that must be preserved
These changes ensure that:
Completed subtasks will be preserved
The AI will build on top of what's already been done
If something needs to be changed/undone, it will be handled through new subtasks
The user is clearly informed about how subtasks are handled.
Fix issue with kebab-case validator incorrectly flagging single-word flags like --prompt. Refactor detectCamelCaseFlags to properly handle all single-word flags. Update tests to verify correct behavior with single-word and camelCase flags. Add support for alternative flag formats in init command (e.g., -my_name). This fixes a bug where users couldn't use the --prompt flag directly and had to use -p instead.
- Fix tests using ES Module best practices instead of complex mocking
- Replace Commander.js mocking with direct action handler testing
- Resolve ES Module import/mock issues and function redeclaration errors
- Fix circular reference issues with console.log spies
- Properly setup mock functions with jest.fn() for method access
- Improve parse-prd command functionality
- Add default PRD path support (scripts/prd.txt) so you can just run `task-master parse-prd` and it will use the default PRD if it exists.
- Improve error handling and user feedback
- Enhance help text with more detailed information
- Fix detectCamelCaseFlags implementation in utils.js yet again with more tests this time
- Improve regex pattern to correctly detect camelCase flags
- Skip flags already in kebab-case format
- Enhance tests with proper test-specific implementations
- Document testing best practices
- Add comprehensive "Common Testing Pitfalls and Solutions" section to tests.mdc
- Provide clear examples of correct testing patterns for ES modules
- Document techniques for test isolation and mock organization
This commit introduces several significant improvements:
- **Enhanced Unit Testing:** Vastly improved unit tests for the module, covering core functions, edge cases, and error handling. Simplified test functions and comprehensive mocking were implemented for better isolation and reliability. Added new section to tests.mdc detailing reliable testing techniques.
- **CLI Kebab-Case Flag Enforcement:** The CLI now enforces kebab-case for flags, providing helpful error messages when camelCase is used. This improves consistency and user experience.
- **AI Enhancements:**
- Enabled 128k token output for Claude 3.7 Sonnet by adding the header.
- Added a new task to to document this change and its testing strategy.
- Added unit tests to verify the Anthropic client configuration.
- Added and utility functions.
- **Improved Test Coverage:** Added tests for the new CLI flag validation logic.
