* initial cutover
* update log to debug
* update tracker to pass units
* update test to match new base tracker format
* add streamTextService mocks
* remove unused imports
* Ensure the CLI waits for async main() completion
* refactor to reduce code duplication
* update comment
* reuse function
* ensure targetTag is defined in streaming mode
* avoid throwing inside process.exit spy
* check for null
* remove reference to generate
* fix formatting
* fix textStream assignment
* ensure no division by 0
* fix jest chalk mocks
* refactor for maintainability
* Improve bar chart calculation logic for consistent visual representation
* use custom streaming error types; fix mocks
* Update streamText extraction in parse-prd.js to match actual service response
* remove check - doesn't belong here
* update mocks
* remove streaming test that wasn't really doing anything
* add comment
* make parsing logic more DRY
* fix formatting
* Fix textStream extraction to match actual service response
* fix mock
* Add a cleanup method to ensure proper resource disposal and prevent memory leaks
* debounce progress updates to reduce UI flicker during rapid updates
* Implement timeout protection for streaming operations (60-second timeout) with automatic fallback to non-streaming mode.
* clear timeout properly
* Add a maximum buffer size limit (1MB) to prevent unbounded memory growth with very large streaming responses.
* fix formatting
* remove duplicate mock
* better docs
* fix formatting
* sanitize the dynamic property name
* Fix incorrect remaining progress calculation
* Use onError callback instead of console.warn
* Remove unused chalk import
* Add missing custom validator in fallback parsing configuration
* add custom validator parameter in fallback parsing
* chore: fix package-lock.json
* chore: large code refactor
* chore: increase timeout from 1 minute to 3 minutes
* fix: refactor and fix streaming
* Merge remote-tracking branch 'origin/next' into joedanz/parse-prd-progress
* fix: cleanup and fix unit tests
* chore: fix unit tests
* chore: fix format
* chore: run format
* chore: fix weird CI unit test error
* chore: fix format
---------
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
* Version Packages
* chore: add eyal instead of crunchyman to eyal's feature
* chore: run format
---------
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
* fix: fix mcp tool call in extension
- fix console.log directly being used in scope-adjutment.js breaking mcp
* chore: run format and fix tests
* chore: format
* Version Packages
* chore: fix release 0.22
todo: fix CI
* chore: run format
---------
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
* fix: bump ai-sdk-provider-gemini-cli to v0.1.1
Updates ai-sdk-provider-gemini-cli from v0.0.4 to v0.1.1 to fix a breaking change
introduced in @google/gemini-cli-core v0.1.12+ where createContentGeneratorConfig
signature changed, causing "config.getModel is not a function" errors.
The new version includes:
- Fixed compatibility with @google/gemini-cli-core ^0.1.13
- Added proxy support via configuration
- Resolved the breaking API change
Fixes compatibility issues when using newer versions of gemini-cli-core.
See: https://github.com/ben-vargas/ai-sdk-provider-gemini-cli/releases/tag/v0.1.1
* chore: fix package-lock.json being too big
---------
Co-authored-by: Ben Vargas <ben@example.com>
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
* fix(profiles): fix vscode profile generation
- Add .instructions.md extension for VSCode Copilot instructions file.
- Add customReplacement to remove unsupported property `alwaysApply` from YAML front-matter in VSCode instructions files.
- Add missing property `targetExtension` to the base profile object to
support the change to file extension.
* chore: run format
---------
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
* move claude rules and commands to assets/claude
* update claude profile to copy assets/claude to .claude
* fix formatting
* feat(profiles): Implement unified profile system
- Convert Claude and Codex profiles to use createProfile() factory
- Remove simple vs complex profile distinction in rule transformer
- Unify convertAllRulesToProfileRules() to handle all profiles consistently
- Fix mcpConfigPath construction in base-profile.js for null mcpConfigName
- Update terminology from 'simpleProfiles' to 'assetOnlyProfiles' throughout
- Ensure Claude .claude directory copying works in both CLI and MCP contexts
- All profiles now follow same execution flow with proper lifecycle functions
Changes:
- src/profiles/claude.js: Convert to createProfile() factory pattern
- src/profiles/codex.js: Convert to createProfile() factory pattern
- src/utils/rule-transformer.js: Unified profile handling logic
- src/utils/profiles.js: Remove simple profile categorization
- src/profiles/base-profile.js: Fix mcpConfigPath construction
- scripts/modules/commands.js: Update variable naming
- tests/: Update all tests for unified system and terminology
Fixes Claude profile asset copying issue in MCP context.
All tests passing (617 passed, 11 skipped).
* re-checkin claude files
* fix formatting
* chore: clean up test Claude rules files
* chore: add changeset for unified profile system
* add claude files back
* add changeset
* restore proper gitignore
* remove claude agents file from root
* remove incorrect doc
* simplify profiles and update tests
* update changeset
* update changeset
* remove profile specific code
* streamline profiles with defaults and update tests
* update changeset
* add newline at end of gitignore
* restore changes
* streamline profiles with defaults; update tests and add vscode test
* update rule profile tests
* update wording for clearer profile management
* refactor and clarify terminology
* use original projectRoot var name
* revert param desc
* use updated claude assets from neno
* add "YOUR_" before api key here
* streamline codex profile
* add gemini profile
* update gemini profile
* update tests
* relocate function
* update rules interactive setup Gemini desc
* remove duplicative code
* add comma
* feat: support MCP sampling
* support provider registry
* use standard config options for MCP provider
* update fastmcp to support passing params to requestSampling
* move key name definition to base provider
* moved check for required api key to provider class
* remove unused code
* more cleanup
* more cleanup
* refactor provider
* remove not needed files
* more cleanup
* more cleanup
* more cleanup
* update docs
* fix tests
* add tests
* format fix
* clean files
* merge fixes
* format fix
* feat: add support for MCP Sampling as AI provider
* initial mcp ai sdk
* fix references to old provider
* update models
* lint
* fix gemini-cli conflicts
* ran format
* Update src/provider-registry/index.js
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
* fix circular dependency
Circular Dependency Issue ✅ FIXED
Root Cause: BaseAIProvider was importing from index.js, which includes commands.js and other modules that eventually import back to AI providers
Solution: Changed imports to use direct paths to avoid circular dependencies:
Updated base-provider.js to import log directly from utils.js
Updated gemini-cli.js to import log directly from utils.js
Result: Fixed 11 failing tests in mcp-provider.test.js
* fix gemini test
* fix(claude-code): recover from CLI JSON truncation bug (#913) (#920)
Gracefully handle SyntaxError thrown by @anthropic-ai/claude-code when the CLI truncates large JSON outputs (4–16 kB cut-offs).\n\nKey points:\n• Detect JSON parse error + existing buffered text in both doGenerate() and doStream() code paths.\n• Convert the failure into a recoverable 'truncated' finish state and push a provider-warning.\n• Allows Task Master to continue parsing long PRDs / expand-task operations instead of crashing.\n\nA patch changeset (.changeset/claude-code-json-truncation.md) is included for the next release.\n\nRef: eyaltoledano/claude-task-master#913
* docs: fix gemini-cli authentication documentation (#923)
Remove erroneous 'gemini auth login' command references and replace with correct 'gemini' command authentication flow. Update documentation to reflect proper OAuth setup process via the gemini CLI interactive interface.
* fix tests
* fix: update ai-sdk-provider-gemini-cli to 0.0.4 for improved authentication (#932)
- Fixed authentication compatibility issues with Google auth
- Added support for 'api-key' auth type alongside 'gemini-api-key'
- Resolved "Unsupported authType: undefined" runtime errors
- Updated @google/gemini-cli-core dependency to 0.1.9
- Improved documentation and removed invalid auth references
- Maintained backward compatibility while enhancing type validation
* call logging directly
Need to patch upstream fastmcp to allow easier access and bootstrap the TM mcp logger to use the fastmcp logger which today is only exposed in the tools handler
* fix tests
* removing logs until we figure out how to pass mcp logger
* format
* fix tests
* format
* clean up
* cleanup
* readme fix
---------
Co-authored-by: Oren Melamed <oren.m@gloat.com>
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
Co-authored-by: Ben Vargas <ben@vargas.com>
* feat: Add GROQ API key support and integrate GROQ provider
* feat: Add support for Groq provider
- Added a new changeset documenting the addition of Groq provider support.
-Ran npm run format
* feat: Add support for Groq provider
- Added a new changeset documenting the addition of Groq provider support.
-Ran npm run format
* feat: Add gemini-cli provider integration for Task Master
This commit adds comprehensive support for the Gemini CLI provider, enabling users
to leverage Google's Gemini models through OAuth authentication via the gemini CLI
tool. This integration provides a seamless experience for users who prefer using
their existing Google account authentication rather than managing API keys.
## Implementation Details
### Provider Class (`src/ai-providers/gemini-cli.js`)
- Created GeminiCliProvider extending BaseAIProvider
- Implements dual authentication support:
- Primary: OAuth authentication via `gemini auth login` (authType: 'oauth-personal')
- Secondary: API key authentication for compatibility (authType: 'api-key')
- Uses the npm package `ai-sdk-provider-gemini-cli` (v0.0.3) for SDK integration
- Properly handles authentication validation without console output
### Model Configuration (`scripts/modules/supported-models.json`)
- Added two Gemini models with accurate specifications:
- gemini-2.5-pro: 72% SWE score, 65,536 max output tokens
- gemini-2.5-flash: 71% SWE score, 65,536 max output tokens
- Both models support main, fallback, and research roles
- Configured with zero cost (free tier)
### System Integration
- Registered provider in PROVIDERS map (`scripts/modules/ai-services-unified.js`)
- Added to OPTIONAL_AUTH_PROVIDERS set for flexible authentication
- Added GEMINI_CLI constant to provider constants (`src/constants/providers.js`)
- Exported GeminiCliProvider from index (`src/ai-providers/index.js`)
### Command Line Support (`scripts/modules/commands.js`)
- Added --gemini-cli flag to models command for provider hint
- Integrated into model selection logic (setModel function)
- Updated error messages to include gemini-cli in provider list
- Removed unrelated azure/vertex changes to maintain PR focus
### Documentation (`docs/providers/gemini-cli.md`)
- Comprehensive provider documentation emphasizing OAuth-first approach
- Clear explanation of why users would choose gemini-cli over standard google provider
- Detailed installation, authentication, and configuration instructions
- Troubleshooting section with common issues and solutions
### Testing (`tests/unit/ai-providers/gemini-cli.test.js`)
- Complete test suite with 12 tests covering all functionality
- Tests for both OAuth and API key authentication paths
- Error handling and edge case coverage
- Updated mocks in ai-services-unified.test.js for integration testing
## Key Design Decisions
1. **OAuth-First Design**: The provider assumes users want to leverage their existing
`gemini auth login` credentials, making this the default authentication method.
2. **Authentication Type Mapping**: Discovered through testing that the SDK expects:
- 'oauth-personal' for OAuth/CLI authentication (not 'gemini-cli' or 'oauth')
- 'api-key' for API key authentication (not 'gemini-api-key')
3. **Silent Operation**: Removed console.log statements from validateAuth to match
the pattern used by other providers like claude-code.
4. **Limited Model Support**: Only gemini-2.5-pro and gemini-2.5-flash are available
through the CLI, as confirmed by the package author.
## Usage
```bash
# Install gemini CLI globally
npm install -g @google/gemini-cli
# Authenticate with Google account
gemini auth login
# Configure Task Master to use gemini-cli
task-master models --set-main gemini-2.5-pro --gemini-cli
# Use Task Master normally
task-master new "Create a REST API endpoint"
```
## Dependencies
- Added `ai-sdk-provider-gemini-cli@^0.0.3` to package.json
- This package wraps the Google Gemini CLI Core functionality for Vercel AI SDK
## Testing
All tests pass (613 total), including the new gemini-cli provider tests.
Code has been formatted with biome to maintain consistency.
This implementation provides a clean, well-tested integration that follows Task Master's
existing patterns while offering users a convenient way to use Gemini models with their
existing Google authentication.
* feat: implement lazy loading for gemini-cli provider
- Move ai-sdk-provider-gemini-cli to optionalDependencies
- Implement dynamic import with loadGeminiCliModule() function
- Make getClient() async to support lazy loading
- Update base-provider to handle async getClient() calls
- Update tests to handle async getClient() method
This allows the application to start without the gemini-cli package
installed, only loading it when actually needed.
* feat(gemini-cli): replace regex-based JSON extraction with jsonc-parser
- Add jsonc-parser dependency for robust JSON parsing
- Replace simple regex approach with progressive parsing strategy:
1. Direct parsing after cleanup
2. Smart boundary detection with single-pass analysis
3. Limited fallback for edge cases
- Optimize performance with early termination and strategic sampling
- Add comprehensive tests for variable declarations, trailing commas,
escaped quotes, nested objects, and performance edge cases
- Improve reliability for complex JSON structures that Gemini commonly produces
- Fix code formatting with biome
This addresses JSON parsing failures in generateObject operations while
maintaining backward compatibility and significantly improving performance
for large responses.
* fix: update package-lock.json and fix formatting for CI/CD
- Add jsonc-parser to package-lock.json for proper npm ci compatibility
- Fix biome formatting issues in gemini-cli provider and tests
- Ensure all CI/CD checks pass
* feat(gemini-cli): implement comprehensive JSON output reliability system
- Add automatic JSON request detection via content analysis patterns
- Implement task-specific prompt simplification for improved AI compliance
- Add strict JSON enforcement through enhanced system prompts
- Implement response interception with intelligent JSON extraction fallback
- Add comprehensive test coverage for all new JSON handling methods
- Move debug logging to appropriate level for clean user experience
This multi-layered approach addresses gemini-cli's conversational response
tendencies, ensuring reliable structured JSON output for task expansion
operations. Achieves 100% success rate in end-to-end testing while
maintaining full backward compatibility with existing functionality.
Technical implementation includes:
• JSON detection via user message content analysis
• Expand-task prompt simplification with cleaner instructions
• System prompt enhancement with strict JSON enforcement
• Response processing with jsonc-parser-based extraction
• Comprehensive unit test coverage for edge cases
• Debug-level logging to prevent user interface clutter
Resolves: gemini-cli JSON formatting inconsistencies
Tested: All 46 test suites pass, formatting verified
* chore: add changeset for gemini-cli provider implementation
Adds minor version bump for comprehensive gemini-cli provider with:
- Lazy loading and optional dependency management
- Advanced JSON parsing with jsonc-parser
- Multi-layer reliability system for structured output
- Complete test coverage and CI/CD compliance
* refactor: consolidate optional auth provider logic
- Add gemini-cli to existing providersWithoutApiKeys array in config-manager
- Export providersWithoutApiKeys for reuse across modules
- Remove duplicate OPTIONAL_AUTH_PROVIDERS Set from ai-services-unified
- Update ai-services-unified to import and use centralized array
- Fix Jest mock to include new providersWithoutApiKeys export
This eliminates code duplication and provides a single source of truth
for which providers support optional authentication, addressing PR
reviewer feedback about existing similar functionality in src/constants.
* Update SWE scores (#657)
* docs: Auto-update and format models.md
* feat: Flexible brand rules management (#460)
* chore(docs): update docs and rules related to model management.
* feat(ai): Add OpenRouter AI provider support
Integrates the OpenRouter AI provider using the Vercel AI SDK adapter (@openrouter/ai-sdk-provider). This allows users to configure and utilize models available through the OpenRouter platform.
- Added src/ai-providers/openrouter.js with standard Vercel AI SDK wrapper functions (generateText, streamText, generateObject).
- Updated ai-services-unified.js to include the OpenRouter provider in the PROVIDER_FUNCTIONS map and API key resolution logic.
- Verified config-manager.js handles OpenRouter API key checks correctly.
- Users can configure OpenRouter models via .taskmasterconfig using the task-master models command or MCP models tool. Requires OPENROUTER_API_KEY.
- Enhanced error handling in ai-services-unified.js to provide clearer messages when generateObjectService fails due to lack of underlying tool support in the selected model/provider endpoint.
* feat(cli): Add --status/-s filter flag to show command and get-task MCP tool
Implements the ability to filter subtasks displayed by the `task-master show <id>` command using the `--status` (or `-s`) flag. This is also available in the MCP context.
- Modified `commands.js` to add the `--status` option to the `show` command definition.
- Updated `utils.js` (`findTaskById`) to handle the filtering logic and return original subtask counts/arrays when filtering.
- Updated `ui.js` (`displayTaskById`) to use the filtered subtasks for the table, display a summary line when filtering, and use the original subtask list for the progress bar calculation.
- Updated MCP `get_task` tool and `showTaskDirect` function to accept and pass the `status` parameter.
- Added changeset entry.
* fix(tasks): Improve next task logic to be subtask-aware
* fix(tasks): Enable removing multiple tasks/subtasks via comma-separated IDs
- Refactors the core `removeTask` function (`task-manager/remove-task.js`) to accept and iterate over comma-separated task/subtask IDs.
- Updates dependency cleanup and file regeneration logic to run once after processing all specified IDs.
- Adjusts the `remove-task` CLI command (`commands.js`) description and confirmation prompt to handle multiple IDs correctly.
- Fixes a bug in the CLI confirmation prompt where task/subtask titles were not being displayed correctly.
- Updates the `remove_task` MCP tool description to reflect the new multi-ID capability.
This addresses the previously known issue where only the first ID in a comma-separated list was processed.
Closes#140
* Update README.md (#342)
* Update Discord badge (#337)
* refactor(init): Improve robustness and dependencies; Update template deps for AI SDKs; Silence npm install in MCP; Improve conditional model setup logic; Refactor init.js flags; Tweak Getting Started text; Fix MCP server launch command; Update default model in config template
* Refactor: Improve MCP logging, update E2E & tests
Refactors MCP server logging and updates testing infrastructure.
- MCP Server:
- Replaced manual logger wrappers with centralized `createLogWrapper` utility.
- Updated direct function calls to use `{ session, mcpLog }` context.
- Removed deprecated `model` parameter from analyze, expand-all, expand-task tools.
- Adjusted MCP tool import paths and parameter descriptions.
- Documentation:
- Modified `docs/configuration.md`.
- Modified `docs/tutorial.md`.
- Testing:
- E2E Script (`run_e2e.sh`):
- Removed `set -e`.
- Added LLM analysis function (`analyze_log_with_llm`) & integration.
- Adjusted test run directory creation timing.
- Added debug echo statements.
- Deleted Unit Tests: Removed `ai-client-factory.test.js`, `ai-client-utils.test.js`, `ai-services.test.js`.
- Modified Fixtures: Updated `scripts/task-complexity-report.json`.
- Dev Scripts:
- Modified `scripts/dev.js`.
* chore(tests): Passes tests for merge candidate
- Adjusted the interactive model default choice to be 'no change' instead of 'cancel setup'
- E2E script has been perfected and works as designed provided there are all provider API keys .env in the root
- Fixes the entire test suite to make sure it passes with the new architecture.
- Fixes dependency command to properly show there is a validation failure if there is one.
- Refactored config-manager.test.js mocking strategy and fixed assertions to read the real supported-models.json
- Fixed rule-transformer.test.js assertion syntax and transformation logic adjusting replacement for search which was too broad.
- Skip unstable tests in utils.test.js (log, readJSON, writeJSON error paths) due to SIGABRT crash. These tests trigger a native crash (SIGABRT), likely stemming from a conflict between internal chalk usage within the functions and Jest's test environment, possibly related to ESM module handling.
* chore(wtf): removes chai. not sure how that even made it in here. also removes duplicate test in scripts/.
* fix: ensure API key detection properly reads .env in MCP context
Problem:
- Task Master model configuration wasn't properly checking for API keys in the project's .env file when running through MCP
- The isApiKeySet function was only checking session.env and process.env but not inspecting the .env file directly
- This caused incorrect API key status reporting in MCP tools even when keys were properly set in .env
Solution:
- Modified resolveEnvVariable function in utils.js to properly read from .env file at projectRoot
- Updated isApiKeySet to correctly pass projectRoot to resolveEnvVariable
- Enhanced the key detection logic to have consistent behavior between CLI and MCP contexts
- Maintains the correct precedence: session.env → .env file → process.env
Testing:
- Verified working correctly with both MCP and CLI tools
- API keys properly detected in .env file in both contexts
- Deleted .cursor/mcp.json to confirm introspection of .env as fallback works
* fix(update): pass projectRoot through update command flow
Modified ai-services-unified.js, update.js tool, and update-tasks.js direct function to correctly pass projectRoot. This enables the .env file API key fallback mechanism for the update command when running via MCP, ensuring consistent key resolution with the CLI context.
* fix(analyze-complexity): pass projectRoot through analyze-complexity flow
Modified analyze-task-complexity.js core function, direct function, and analyze.js tool to correctly pass projectRoot. Fixed import error in tools/index.js. Added debug logging to _resolveApiKey in ai-services-unified.js. This enables the .env API key fallback for analyze_project_complexity.
* fix(add-task): pass projectRoot and fix logging/refs
Modified add-task core, direct function, and tool to pass projectRoot for .env API key fallback. Fixed logFn reference error and removed deprecated reportProgress call in core addTask function. Verified working.
* fix(parse-prd): pass projectRoot and fix schema/logging
Modified parse-prd core, direct function, and tool to pass projectRoot for .env API key fallback. Corrected Zod schema used in generateObjectService call. Fixed logFn reference error in core parsePRD. Updated unit test mock for utils.js.
* fix(update-task): pass projectRoot and adjust parsing
Modified update-task-by-id core, direct function, and tool to pass projectRoot. Reverted parsing logic in core function to prioritize `{...}` extraction, resolving parsing errors. Fixed ReferenceError by correctly destructuring projectRoot.
* fix(update-subtask): pass projectRoot and allow updating done subtasks
Modified update-subtask-by-id core, direct function, and tool to pass projectRoot for .env API key fallback. Removed check preventing appending details to completed subtasks.
* fix(mcp, expand): pass projectRoot through expand/expand-all flows
Problem: expand_task & expand_all MCP tools failed with .env keys due to missing projectRoot propagation for API key resolution. Also fixed a ReferenceError: wasSilent is not defined in expandTaskDirect.
Solution: Modified core logic, direct functions, and MCP tools for expand-task and expand-all to correctly destructure projectRoot from arguments and pass it down through the context object to the AI service call (generateTextService). Fixed wasSilent scope in expandTaskDirect.
Verification: Tested expand_task successfully in MCP using .env keys. Reviewed expand_all flow for correct projectRoot propagation.
* chore: prettier
* fix(expand-all): add projectRoot to expandAllTasksDirect invokation.
* fix(update-tasks): Improve AI response parsing for 'update' command
Refactors the JSON array parsing logic within
in .
The previous logic primarily relied on extracting content from markdown
code blocks (json or javascript), which proved brittle when the AI
response included comments or non-JSON text within the block, leading to
parsing errors for the command.
This change modifies the parsing strategy to first attempt extracting
content directly between the outermost '[' and ']' brackets. This is
more robust as it targets the expected array structure directly. If
bracket extraction fails, it falls back to looking for a strict json
code block, then prefix stripping, before attempting a raw parse.
This approach aligns with the successful parsing strategy used for
single-object responses in and resolves the
parsing errors previously observed with the command.
* refactor(mcp): introduce withNormalizedProjectRoot HOF for path normalization
Added HOF to mcp tools utils to normalize projectRoot from args/session. Refactored get-task tool to use HOF. Updated relevant documentation.
* refactor(mcp): apply withNormalizedProjectRoot HOF to update tool
Problem: The MCP tool previously handled project root acquisition and path resolution within its method, leading to potential inconsistencies and repetition.
Solution: Refactored the tool () to utilize the new Higher-Order Function (HOF) from .
Specific Changes:
- Imported HOF.
- Updated the Zod schema for the parameter to be optional, as the HOF handles deriving it from the session if not provided.
- Wrapped the entire function body with the HOF.
- Removed the manual call to from within the function body.
- Destructured the from the object received by the wrapped function, ensuring it's the normalized path provided by the HOF.
- Used the normalized variable when calling and when passing arguments to .
This change standardizes project root handling for the tool, simplifies its method, and ensures consistent path normalization. This serves as the pattern for refactoring other MCP tools.
* fix: apply to all tools withNormalizedProjectRoot to fix projectRoot issues for linux and windows
* fix: add rest of tools that need wrapper
* chore: cleanup tools to stop using rootFolder and remove unused imports
* chore: more cleanup
* refactor: Improve update-subtask, consolidate utils, update config
This commit introduces several improvements and refactorings across MCP tools, core logic, and configuration.
**Major Changes:**
1. **Refactor updateSubtaskById:**
- Switched from generateTextService to generateObjectService for structured AI responses, using a Zod schema (subtaskSchema) for validation.
- Revised prompts to have the AI generate relevant content based on user request and context (parent/sibling tasks), while explicitly preventing AI from handling timestamp/tag formatting.
- Implemented **local timestamp generation (new Date().toISOString()) and formatting** (using <info added on ...> tags) within the function *after* receiving the AI response. This ensures reliable and correctly formatted details are appended.
- Corrected logic to append only the locally formatted, AI-generated content block to the existing subtask.details.
2. **Consolidate MCP Utilities:**
- Moved/consolidated the withNormalizedProjectRoot HOF into mcp-server/src/tools/utils.js.
- Updated MCP tools (like update-subtask.js) to import withNormalizedProjectRoot from the new location.
3. **Refactor Project Initialization:**
- Deleted the redundant mcp-server/src/core/direct-functions/initialize-project-direct.js file.
- Updated mcp-server/src/core/task-master-core.js to import initializeProjectDirect from its correct location (./direct-functions/initialize-project.js).
**Other Changes:**
- Updated .taskmasterconfig fallback model to claude-3-7-sonnet-20250219.
- Clarified model cost representation in the models tool description (taskmaster.mdc and mcp-server/src/tools/models.js).
* fix: displayBanner logging when silentMode is active (#385)
* fix: improve error handling, test options, and model configuration
- Enhance error validation in parse-prd.js and update-tasks.js
- Fix bug where mcpLog was incorrectly passed as logWrapper
- Improve error messages and response formatting
- Add --skip-verification flag to E2E tests
- Update MCP server config that ships with init to match new API key structure
- Fix task force/append handling in parse-prd command
- Increase column width in update-tasks display
* chore: fixes parse prd to show loading indicator in cli.
* fix(parse-prd): suggested fix for mcpLog was incorrect. reverting to my previously working code.
* chore(init): No longer ships readme with task-master init (commented out for now). No longer looking for task-master-mcp, instead checked for task-master-ai - this should prevent the init sequence from needlessly adding another mcp server with task-master-mcp to the mpc.json which a ton of people probably ran into.
* chore: restores 3.7 sonnet as the main role.
* fix(add/remove-dependency): dependency mcp tools were failing due to hard-coded tasks path in generate task files.
* chore: removes tasks json backup that was temporarily created.
* fix(next): adjusts mcp tool response to correctly return the next task/subtask. Also adds nextSteps to the next task response.
* chore: prettier
* chore: readme typos
* fix(config): restores sonnet 3.7 as default main role.
* Version Packages
* hotfix: move production package to "dependencies" (#399)
* Version Packages
* Fix: issues with 0.13.0 not working (#402)
* Exit prerelease mode and version packages
* hotfix: move production package to "dependencies"
* Enter prerelease mode and version packages
* Enter prerelease mode and version packages
* chore: cleanup
* chore: improve pre.json and add pre-release workflow
* chore: fix package.json
* chore: cleanup
* chore: improve pre-release workflow
* chore: allow github actions to commit
* extract fileMap and conversionConfig into brand profile
* extract into brand profile
* add windsurf profile
* add remove brand rules function
* fix regex
* add rules command to add/remove rules for a specific brand
* fix post processing for roo
* allow multiples
* add cursor profile
* update test for new structure
* move rules to assets
* use assets/rules for rules files
* use standardized setupMCP function
* fix formatting
* fix formatting
* add logging
* fix escapes
* default to cursor
* allow init with certain rulesets; no more .windsurfrules
* update docs
* update log msg
* fix formatting
* keep mdc extension for cursor
* don't rewrite .mdc to .md inside the files
* fix roo init (add modes)
* fix cursor init (don't use roo transformation by default)
* use more generic function names
* update docs
* fix formatting
* update function names
* add changeset
* add rules to mcp initialize project
* register tool with mcp server
* update docs
* add integration test
* fix cursor initialization
* rule selection
* fix formatting
* fix MCP - remove yes flag
* add import
* update roo tests
* add/update tests
* remove test
* add rules command test
* update MCP responses, centralize rules profiles & helpers
* fix logging and MCP response messages
* fix formatting
* incorrect test
* fix tests
* update fileMap
* fix file extension transformations
* fix formatting
* add rules command test
* test already covered
* fix formatting
* move renaming logic into profiles
* make sure dir is deleted (DS_Store)
* add confirmation for rules removal
* add force flag for rules remove
* use force flag for test
* remove yes parameter
* fix formatting
* import brand profiles from rule-transformer.js
* update comment
* add interactive rules setup
* optimize
* only copy rules specifically listed in fileMap
* update comment
* add cline profile
* add brandDir to remove ambiguity and support Cline
* specify whether to create mcp config and filename
* add mcpConfigName value for parh
* fix formatting
* remove rules just for this repository - only include rules to be distributed
* update error message
* update "brand rules" to "rules"
* update to minor
* remove comment
* remove comments
* move to /src/utils
* optimize imports
* move rules-setup.js to /src/utils
* move rule-transformer.js to /src/utils
* move confirmation to /src/ui/confirm.js
* default to all rules
* use profile js for mcp config settings
* only run rules interactive setup if not provided via command line
* update comments
* initialize with all brands if nothing specified
* update var name
* clean up
* enumerate brands for brand rules
* update instructions
* add test to check for brand profiles
* fix quotes
* update semantics and terminology from 'brand rules' to 'rules profiles'
* fix formatting
* fix formatting
* update function name and remove copying of cursor rules, now handled by rules transformer
* update comment
* rename to mcp-config-setup.js
* use enums for rules actions
* add aggregate reporting for rules add command
* add missing log message
* use simpler path
* use base profile with modifications for each brand
* use displayName and don't select any defaults in setup
* add confirmation if removing ALL rules profiles, and add --force flag on rules remove
* Use profile-detection instead of rules-detection
* add newline at end of mcp config
* add proper formatting for mcp.json
* update rules
* update rules
* update rules
* add checks for other rules and other profile folder items before removing
* update confirmation for rules remove
* update docs
* update changeset
* fix for filepath at bottom of rule
* Update cline profile and add test; adjust other rules tests
* update changeset
* update changeset
* clarify init for all profiles if not specified
* update rule text
* revert text
* use "rule profiles" instead of "rules profiles"
* use standard tool mappings for windsurf
* add Trae support
* update changeset
* update wording
* update to 'rule profile'
* remove unneeded exports to optimize loc
* combine to /src/utils/profiles.js; add codex and claude code profiles
* rename function and add boxen
* add claude and codex integration tests
* organize tests into profiles folder
* mock fs for transformer tests
* update UI
* add cline and trae integration tests
* update test
* update function name
* update formatting
* Update change set with new profiles
* move profile integration tests to subdirectory
* properly create temp directories in /tmp folder
* fix formatting
* use taskmaster subfolder for the 2 TM rules
* update wording
* ensure subdirectory exists
* update rules from next
* update from next
* update taskmaster rule
* add details on new rules command and init
* fix mcp init
* fix MCP path to assets
* remove duplication
* remove duplication
* MCP server path fixes for rules command
* fix for CLI roo rules add/remove
* update tests
* fix formatting
* fix pattern for interactive rule profiles setup
* restore comments
* restore comments
* restore comments
* remove unused import, fix quotes
* add missing integration tests
* add VS Code profile and tests
* update docs and rules to include vscode profile
* add rules subdirectory support per-profile
* move profiles to /src
* fix formatting
* rename to remove ambiguity
* use --setup for rules interactive setup
* Fix Cursor deeplink installation with copy-paste instructions (#723)
* change roo boomerang to orchestrator; update tests that don't use modes
* fix newline
* chore: cleanup
---------
Co-authored-by: Eyal Toledano <eyal@microangel.so>
Co-authored-by: Yuval <yuvalbl@users.noreply.github.com>
Co-authored-by: Marijn van der Werf <marijn.vanderwerf@gmail.com>
Co-authored-by: Eyal Toledano <eutait@gmail.com>
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* fix: providers config for azure, bedrock, and vertex (#822)
* fix: providers config for azure, bedrock, and vertex
* chore: improve changelog
* chore: fix CI
* fix: switch to ESM export to avoid mixed format (#633)
* fix: switch to ESM export to avoid mixed format
The CLI entrypoint was using `module.exports` alongside ESM `import` statements,
resulting in an invalid mixed module format. Replaced the CommonJS export with
a proper ESM `export` to maintain consistency and prevent module resolution issues.
* chore: add changeset
---------
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
* fix: Fix external provider support (#726)
* fix(bedrock): improve AWS credential handling and add model definitions (#826)
* fix(bedrock): improve AWS credential handling and add model definitions
- Change error to warning when AWS credentials are missing in environment
- Allow fallback to system configuration (aws config files or instance profiles)
- Remove hardcoded region and profile parameters in Bedrock client
- Add Claude 3.7 Sonnet and DeepSeek R1 model definitions for Bedrock
- Update config manager to properly handle Bedrock provider
* chore: cleanup and format and small refactor
---------
Co-authored-by: Ray Krueger <raykrueger@gmail.com>
* docs: Auto-update and format models.md
* Version Packages
* chore: fix package.json
* Fix/expand command tag corruption (#827)
* fix(expand): Fix tag corruption in expand command - Fix tag parameter passing through MCP expand-task flow - Add tag parameter to direct function and tool registration - Fix contextGatherer method name from _buildDependencyContext to _buildDependencyGraphs - Add comprehensive test coverage for tag handling in expand-task - Ensures tagged task structure is preserved during expansion - Prevents corruption when tag is undefined. Fixes expand command causing tag corruption in tagged task lists. All existing tests pass and new test coverage added.
* test(e2e): Add comprehensive tag-aware expand testing to verify tag corruption fix - Add new test section for feature-expand tag creation and testing - Verify tag preservation during expand, force expand, and expand --all operations - Test that master tag remains intact and feature-expand tag receives subtasks correctly - Fix file path references to use correct .taskmaster/tasks/tasks.json location - Fix config file check to use .taskmaster/config.json instead of .taskmasterconfig - All tag corruption verification tests pass successfully in E2E test
* fix(changeset): Update E2E test improvements changeset to properly reflect tag corruption fix verification
* chore(changeset): combine duplicate changesets for expand tag corruption fix
Merge eighty-breads-wonder.md into bright-llamas-enter.md to consolidate
the expand command fix and its comprehensive E2E testing enhancements
into a single changeset entry.
* Delete .changeset/eighty-breads-wonder.md
* Version Packages
* chore: fix package.json
* fix(expand): Enhance context handling in expandAllTasks function
- Added `tag` to context destructuring for better context management.
- Updated `readJSON` call to include `contextTag` for improved data integrity.
- Ensured the correct tag is passed during task expansion to prevent tag corruption.
---------
Co-authored-by: Parththipan Thaniperumkarunai <parththipan.thaniperumkarunai@milkmonkey.de>
Co-authored-by: Parthy <52548018+mm-parthy@users.noreply.github.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* Add pyproject.toml as project root marker (#804)
* feat: Add pyproject.toml as project root marker - Added 'pyproject.toml' to the project markers array in findProjectRoot() - Enables Task Master to recognize Python projects using pyproject.toml - Improves project root detection for modern Python development workflows - Maintains compatibility with existing Node.js and Git-based detection
* chore: add changeset
---------
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
* feat: add Claude Code provider support
Implements Claude Code as a new AI provider that uses the Claude Code CLI
without requiring API keys. This enables users to leverage Claude models
through their local Claude Code installation.
Key changes:
- Add complete AI SDK v1 implementation for Claude Code provider
- Custom SDK with streaming/non-streaming support
- Session management for conversation continuity
- JSON extraction for object generation mode
- Support for advanced settings (maxTurns, allowedTools, etc.)
- Integrate Claude Code into Task Master's provider system
- Update ai-services-unified.js to handle keyless authentication
- Add provider to supported-models.json with opus/sonnet models
- Ensure correct maxTokens values are applied (opus: 32000, sonnet: 64000)
- Fix maxTokens configuration issue
- Add max_tokens property to getAvailableModels() output
- Update setModel() to properly handle claude-code models
- Create update-config-tokens.js utility for init process
- Add comprehensive documentation
- User guide with configuration examples
- Advanced settings explanation and future integration options
The implementation maintains full backward compatibility with existing
providers while adding seamless Claude Code support to all Task Master
commands.
* fix(docs): correct invalid commands in claude-code usage examples
- Remove non-existent 'do', 'estimate', and 'analyze' commands
- Replace with actual Task Master commands: next, show, set-status
- Use correct syntax for parse-prd and analyze-complexity
* feat: make @anthropic-ai/claude-code an optional dependency
This change makes the Claude Code SDK package optional, preventing installation failures for users who don't need Claude Code functionality.
Changes:
- Added @anthropic-ai/claude-code to optionalDependencies in package.json
- Implemented lazy loading in language-model.js to only import the SDK when actually used
- Updated documentation to explain the optional installation requirement
- Applied formatting fixes to ensure code consistency
Benefits:
- Users without Claude Code subscriptions don't need to install the dependency
- Reduces package size for users who don't use Claude Code
- Prevents installation failures if the package is unavailable
- Provides clear error messages when the package is needed but not installed
The implementation uses dynamic imports to load the SDK only when doGenerate() or doStream() is called, ensuring the provider can be instantiated without the package present.
* test: add comprehensive tests for ClaudeCodeProvider
Addresses code review feedback about missing automated tests for the ClaudeCodeProvider.
## Changes
- Added unit tests for ClaudeCodeProvider class covering constructor, validateAuth, and getClient methods
- Added unit tests for ClaudeCodeLanguageModel testing lazy loading behavior and error handling
- Added integration tests verifying optional dependency behavior when @anthropic-ai/claude-code is not installed
## Test Coverage
1. **Unit Tests**:
- ClaudeCodeProvider: Basic functionality, no API key requirement, client creation
- ClaudeCodeLanguageModel: Model initialization, lazy loading, error messages, warning generation
2. **Integration Tests**:
- Optional dependency behavior when package is not installed
- Clear error messages for users about missing package
- Provider instantiation works but usage fails gracefully
All tests pass and provide comprehensive coverage for the claude-code provider implementation.
* revert: remove maxTokens update functionality from init
This functionality was out of scope for the Claude Code provider PR.
The automatic updating of maxTokens values in config.json during
initialization is a general improvement that should be in a separate PR.
Additionally, Claude Code ignores maxTokens and temperature parameters
anyway, making this change irrelevant for the Claude Code integration.
Removed:
- scripts/modules/update-config-tokens.js
- Import and usage in scripts/init.js
* docs: add Claude Code support information to README
- Added Claude Code to the list of supported providers in Requirements section
- Noted that Claude Code requires no API key but needs Claude Code CLI
- Added example of configuring claude-code/sonnet model
- Created dedicated Claude Code Support section with key information
- Added link to detailed Claude Code setup documentation
This ensures users are aware of the Claude Code option as a no-API-key
alternative for using Claude models.
* style: apply biome formatting to test files
* fix(models): add missing --claude-code flag to models command
The models command was missing the --claude-code provider flag, preventing users from setting Claude Code models via CLI. While the backend already supported claude-code as a provider hint, there was no command-line flag to trigger it.
Changes:
- Added --claude-code option to models command alongside existing provider flags
- Updated provider flags validation to include claudeCode option
- Added claude-code to providerHint logic for all three model roles (main, research, fallback)
- Updated error message to include --claude-code in list of mutually exclusive flags
- Added example usage in help text
This allows users to properly set Claude Code models using commands like:
task-master models --set-main sonnet --claude-code
task-master models --set-main opus --claude-code
Without this flag, users would get "Model ID not found" errors when trying to set claude-code models, as the system couldn't determine the correct provider for generic model names like "sonnet" or "opus".
* chore: add changeset for Claude Code provider feature
* docs: Auto-update and format models.md
* readme: add troubleshooting note for MCP tools not working
* Feature/compatibleapisupport (#830)
* add compatible platform api support
* Adjust the code according to the suggestions
* Fully revised as requested: restored all required checks, improved compatibility, and converted all comments to English.
* feat: Add support for compatible API endpoints via baseURL
* chore: Add changeset for compatible API support
* chore: cleanup
* chore: improve changeset
* fix: package-lock.json
* fix: package-lock.json
---------
Co-authored-by: He-Xun <1226807142@qq.com>
* Rename Roo Code "Boomerang" role to "Orchestrator" (#831)
* feat: Enhanced project initialization with Git worktree detection (#743)
* Fix Cursor deeplink installation with copy-paste instructions (#723)
* detect git worktree
* add changeset
* add aliases and git flags
* add changeset
* rename and update test
* add store tasks in git functionality
* update changeset
* fix newline
* remove unused import
* update command wording
* update command option text
* fix: update task by id (#834)
* store tasks in git by default (#835)
* Call rules interactive setup during init (#833)
* chore: rc version bump
* feat: Claude Code slash commands for Task Master (#774)
* Fix Cursor deeplink installation with copy-paste instructions (#723)
* fix: expand-task (#755)
* docs: Update o3 model price (#751)
* docs: Auto-update and format models.md
* docs: Auto-update and format models.md
* feat: Add Claude Code task master commands
Adds Task Master slash commands for Claude Code under /project:tm/ namespace
---------
Co-authored-by: Joe Danziger <joe@ticc.net>
Co-authored-by: Ralph Khreish <35776126+Crunchyman-ralph@users.noreply.github.com>
Co-authored-by: Volodymyr Zahorniak <7808206+zahorniak@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: neno-is-ooo <204701868+neno-is-ooo@users.noreply.github.com>
* feat: make more compatible with "o" family models (#839)
* docs: Auto-update and format models.md
* docs: Add comprehensive Azure OpenAI configuration documentation (#837)
* docs: Add comprehensive Azure OpenAI configuration documentation
- Add detailed Azure OpenAI configuration section with prerequisites, authentication, and setup options
- Include both global and per-model baseURL configuration examples
- Add comprehensive troubleshooting guide for common Azure OpenAI issues
- Update environment variables section with Azure OpenAI examples
- Add Azure OpenAI models to all model tables (Main, Research, Fallback)
- Include prominent Azure configuration example in main documentation
- Fix azureBaseURL format to use correct Azure OpenAI endpoint structure
Addresses common Azure OpenAI setup challenges and provides clear guidance for new users.
* refactor: Move Azure models from docs/models.md to scripts/modules/supported-models.json
- Remove Azure model entries from documentation tables
- Add Azure provider section to supported-models.json with gpt-4o, gpt-4o-mini, and gpt-4-1
- Maintain consistency with existing model configuration structure
* docs: Auto-update and format models.md
* Version Packages
* chore: format fix
---------
Co-authored-by: Riccardo (Ricky) Esclapon <32306488+ries9112@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Joe Danziger <joe@ticc.net>
Co-authored-by: Eyal Toledano <eyal@microangel.so>
Co-authored-by: Yuval <yuvalbl@users.noreply.github.com>
Co-authored-by: Marijn van der Werf <marijn.vanderwerf@gmail.com>
Co-authored-by: Eyal Toledano <eutait@gmail.com>
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Nathan Marley <nathan@glowberrylabs.com>
Co-authored-by: Ray Krueger <raykrueger@gmail.com>
Co-authored-by: Parththipan Thaniperumkarunai <parththipan.thaniperumkarunai@milkmonkey.de>
Co-authored-by: Parthy <52548018+mm-parthy@users.noreply.github.com>
Co-authored-by: ejones40 <ethan.jones@fortyau.com>
Co-authored-by: Ben Vargas <ben@vargas.com>
Co-authored-by: V4G4X <34249137+V4G4X@users.noreply.github.com>
Co-authored-by: He-Xun <1226807142@qq.com>
Co-authored-by: neno <github@meaning.systems>
Co-authored-by: Volodymyr Zahorniak <7808206+zahorniak@users.noreply.github.com>
Co-authored-by: neno-is-ooo <204701868+neno-is-ooo@users.noreply.github.com>
Co-authored-by: Jitesh Thakur <56656484+Jitha-afk@users.noreply.github.com>
This change makes the Claude Code SDK package optional, preventing installation failures for users who don't need Claude Code functionality.
Changes:
- Added @anthropic-ai/claude-code to optionalDependencies in package.json
- Implemented lazy loading in language-model.js to only import the SDK when actually used
- Updated documentation to explain the optional installation requirement
- Applied formatting fixes to ensure code consistency
Benefits:
- Users without Claude Code subscriptions don't need to install the dependency
- Reduces package size for users who don't use Claude Code
- Prevents installation failures if the package is unavailable
- Provides clear error messages when the package is needed but not installed
The implementation uses dynamic imports to load the SDK only when doGenerate() or doStream() is called, ensuring the provider can be instantiated without the package present.
