task-master

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Task Master AI

An AI-powered task management system that integrates seamlessly with AI Agents to manage specification-driven development workflows.

一款基于AI的任务管理系统，可与AI Agent无缝集成，用于管理规范驱动的开发工作流。

Quick Start

快速开始

Three Ways to Use Task Master:

MCP Tools (Recommended) - Direct integration via Model Context Protocol
CLI Commands - Terminal-based task management
Tagged Contexts - Multi-branch/feature task isolation

使用Task Master的三种方式：

MCP Tools（推荐）- 通过Model Context Protocol直接集成
CLI Commands - 基于终端的任务管理
Tagged Contexts - 多分支/功能任务隔离

Core Capabilities

核心功能

Task Management

任务管理

Parse PRDs into actionable tasks automatically
Break down complex tasks into manageable subtasks
Track task dependencies and status
Support for multiple task contexts (tags) for features/branches

自动将PRD解析为可执行任务
将复杂任务拆解为可管理的子任务
追踪任务依赖关系和状态
支持为功能/分支设置多任务上下文（标签）

AI-Powered Features

AI增强功能

Complexity analysis with recommendations
Research-backed task expansion
Intelligent task updates based on implementation drift
Fresh information gathering beyond knowledge cutoff

复杂度分析并提供建议
基于调研的任务扩展
根据实现偏差智能更新任务
获取知识 cutoff 之外的最新信息

Development Workflow

开发工作流

Specification-driven development (SDD) support
Iterative subtask implementation logging
Git branch-aligned task contexts
Team collaboration with isolated task lists

支持规范驱动开发（SDD）
迭代式子任务实现日志记录
与Git分支对齐的任务上下文
支持团队协作的隔离任务列表

When to Use This Skill

何时使用该技能

✅ Use Task Master when:

Starting a new project from a PRD
Managing complex multi-step features
Working on feature branches with isolated tasks
Need to track task dependencies and priorities
Want AI-assisted task breakdown and planning
Collaborating with team members on shared codebase
Need to log implementation progress iteratively

❌ Skip Task Master for:

Simple single-file changes
Quick bug fixes
Trivial tasks with no dependencies
Projects without formal requirements

✅ 推荐使用Task Master的场景：

从PRD启动新项目
管理复杂的多步骤功能
在功能分支上处理隔离的任务
需要追踪任务依赖关系和优先级
希望借助AI辅助进行任务拆解和规划
与团队成员协作开发共享代码库
需要迭代记录实现进度

❌ 不推荐使用Task Master的场景：

简单的单文件修改
快速修复Bug
无依赖的琐碎任务
无正式需求的项目

Setup

设置

Prerequisites

前置条件

Node.js installed
API keys for AI providers (Anthropic, Perplexity, etc.)
Git repository (optional, for branch-based workflows)

已安装Node.js
AI提供商的API密钥（Anthropic、Perplexity等）
Git仓库（可选，用于基于分支的工作流）

Installation

安装

Global Installation:

bash

npm install -g task-master-ai

Project-Local:

bash

npm install task-master-ai

全局安装：

bash

npm install -g task-master-ai

项目本地安装：

bash

npm install task-master-ai

MCP Configuration

MCP配置

Add to your MCP config file (

.cursor/mcp.json

.vscode/mcp.json

, etc.):

json

{
  "mcpServers": {
    "task-master-ai": {
      "command": "npx",
      "args": ["-y", "task-master-ai"],
      "env": {
        "ANTHROPIC_API_KEY": "YOUR_KEY_HERE",
        "PERPLEXITY_API_KEY": "YOUR_KEY_HERE"
      }
    }
  }
}

For complete setup details, see

references/SETUP.md

将以下内容添加到你的MCP配置文件（

.cursor/mcp.json

、

.vscode/mcp.json

等）：

json

{
  "mcpServers": {
    "task-master-ai": {
      "command": "npx",
      "args": ["-y", "task-master-ai"],
      "env": {
        "ANTHROPIC_API_KEY": "YOUR_KEY_HERE",
        "PERPLEXITY_API_KEY": "YOUR_KEY_HERE"
      }
    }
  }
}

完整设置详情请查看

references/SETUP.md

。

Basic Workflow

基础工作流

1. Initialize Project

1. 初始化项目

sql

Initialize taskmaster-ai in my project

sql

Initialize taskmaster-ai in my project

2. Create PRD

2. 创建PRD

Create your Product Requirements Document at

.taskmaster/docs/prd.txt

在

.taskmaster/docs/prd.txt

路径下创建你的产品需求文档（PRD）

3. Parse PRD

3. 解析PRD

sql

Parse my PRD at .taskmaster/docs/prd.txt

sql

Parse my PRD at .taskmaster/docs/prd.txt

4. View Tasks

4. 查看任务

sql

Show me the task list

sql

Show me the task list

5. Work on Tasks

5. 处理任务

sql

What's the next task I should work on?
Can you help me implement task 3?

sql

What's the next task I should work on?
Can you help me implement task 3?

6. Track Progress

6. 追踪进度

sql

Mark task 3 as done
Update subtask 3.2 with my implementation findings

sql

Mark task 3 as done
Update subtask 3.2 with my implementation findings

Key Concepts

核心概念

Tagged Task Lists

带标签的任务列表

Organize tasks into separate contexts (tags) for:

Feature branches (
```
feature-auth
```
,
```
feature-dashboard
```
)
Experiments (
```
experiment-zustand
```
)
Team collaboration (
```
alice-work
```
,
```
bob-work
```
)
Versions (
```
v1.0
```
,
```
v2.0
```
,
```
mvp
```
)

将任务组织到不同的上下文（标签）中，适用于：

功能分支（
```
feature-auth
```
、
```
feature-dashboard
```
）
实验项目（
```
experiment-zustand
```
）
团队协作（
```
alice-work
```
、
```
bob-work
```
）
版本迭代（
```
v1.0
```
、
```
v2.0
```
、
```
mvp
```
）

Task Structure

任务结构

ID: Unique identifier (e.g.,
```
1
```
,
```
1.2
```
)
Title: Brief description
Description: What needs to be done
Status:
```
pending
```
,
```
in-progress
```
,
```
done
```
,
```
deferred
```
Dependencies: Prerequisites (e.g.,
```
[1, 2.1]
```
)
Priority:
```
high
```
,
```
medium
```
,
```
low
```
Details: Implementation notes
Subtasks: Breakdown of complex tasks

ID：唯一标识符（例如：
```
1
```
、
```
1.2
```
）
Title：简短描述
Description：任务内容说明
Status：
```
pending
```
（待处理）、
```
in-progress
```
（进行中）、
```
done
```
（已完成）、
```
deferred
```
（已推迟）
Dependencies：前置任务（例如：
```
[1, 2.1]
```
）
Priority：
```
high
```
（高）、
```
medium
```
（中）、
```
low
```
（低）
Details：实现说明
Subtasks：复杂任务的拆解项

Complexity Analysis

复杂度分析

AI analyzes task complexity (1-10 scale) and recommends:

Number of subtasks needed
Areas requiring research
Implementation approach

AI会分析任务复杂度（1-10分制）并提供以下建议：

需要拆解的子任务数量
需要调研的领域
实现方案

Common Commands

常用命令

Task Viewing

任务查看

sql

List all tasks
Show me task 5
Show me tasks 1, 3, and 5
What's the next task?

sql

List all tasks
Show me task 5
Show me tasks 1, 3, and 5
What's the next task?

Task Creation & Modification

任务创建与修改

sql

Add a task to implement user authentication
Expand task 4 into subtasks
Update task 5 with new requirements
Mark task 3 as done

sql

Add a task to implement user authentication
Expand task 4 into subtasks
Update task 5 with new requirements
Mark task 3 as done

Task Organization

任务组织

sql

Move task 5 to become subtask 7.3
Add dependency: task 8 depends on task 5
Create a new tag called feature-auth
Switch to the feature-auth tag

sql

Move task 5 to become subtask 7.3
Add dependency: task 8 depends on task 5
Create a new tag called feature-auth
Switch to the feature-auth tag

Research & Analysis

调研与分析

sql

Research the latest best practices for JWT authentication
Analyze task complexity for all pending tasks
Expand all pending tasks based on complexity

sql

Research the latest best practices for JWT authentication
Analyze task complexity for all pending tasks
Expand all pending tasks based on complexity

Advanced Workflows

高级工作流

PRD-Driven Feature Development

基于PRD的功能开发

Create dedicated tag for feature
Write comprehensive PRD
Parse PRD into tag
Analyze complexity
Expand complex tasks
Implement iteratively

为功能创建专属标签
编写完整的PRD
将PRD解析到对应标签下
进行复杂度分析
扩展复杂任务
迭代实现

Team Collaboration

团队协作

Create personal tag for your work
Copy tasks from master
Work in isolation
Merge back when ready

为个人工作创建专属标签
从主任务列表复制任务
独立处理任务
完成后合并回主列表

Branch-Based Development

基于分支的开发

Create git branch
Create matching tag from branch
Develop feature with isolated tasks
Merge code and tasks together

创建Git分支
根据分支创建匹配的标签
在隔离的任务列表中开发功能
合并代码与任务

Integration with Development

与开发流程的集成

Iterative Implementation

迭代实现

View subtask details
Plan implementation approach
Log plan to subtask
Begin coding
Log progress and findings
Mark complete
Commit changes

查看子任务详情
规划实现方案
将方案记录到子任务中
开始编码
记录进度和发现
标记任务完成
提交代码变更

Specification-Driven Development

规范驱动开发

Task Master supports full SDD workflow:

Requirements gathering
PRD creation
Task generation
Complexity analysis
Implementation tracking
Progress documentation

Task Master支持完整的SDD工作流：

需求收集
PRD创建
任务生成
复杂度分析
实现追踪
进度文档记录

Resources

资源

```
references/SETUP.md
```
- Complete installation and configuration
```
references/WORKFLOW.md
```
- Detailed development workflows
```
references/COMMANDS.md
```
- Comprehensive command reference
```
references/BEST_PRACTICES.md
```
- Tips and patterns

```
references/SETUP.md
```
- 完整安装与配置指南
```
references/WORKFLOW.md
```
- 详细开发工作流说明
```
references/COMMANDS.md
```
- 全面命令参考
```
references/BEST_PRACTICES.md
```
- 使用技巧与模式

References

参考链接

Quick Tips:

Always start with a detailed PRD
Use complexity analysis before expanding tasks
Log implementation findings to subtasks
Leverage tags for feature isolation
Use research tool for fresh information

快速提示：

始终从详细的PRD开始
在扩展任务前先进行复杂度分析
将实现发现记录到子任务中
利用标签隔离功能开发
使用调研工具获取最新信息