anyt 0.1.1

AI-native task management CLI for coding agents and human developers

AnyTask CLI

AI-native task management CLI for coding agents and human developers

AnyTask CLI is a command-line interface for managing tasks and projects, designed specifically for AI agents and developers working together. It provides Linear-style task management with agent-aware features, MCP (Model Context Protocol) integration, and powerful AI-assisted commands.

Features

Core Capabilities

• Linear-style Task Management: Human-readable task IDs (DEV-123), status workflows, priorities, and dependencies
• Multi-Environment Support: Manage tasks across dev, staging, and production environments
• Workspace & Project Organization: Isolated workspaces with role-based access
• Rich Terminal UI: Beautiful, color-coded task boards and visualizations
• Agent Integration: Native support for AI agents via API keys and MCP server

CLI Commands

• Environment Management: anyt env - Configure and switch between environments
• Authentication: anyt auth login - Authenticate with user tokens or agent API keys
• Workspace Management: anyt workspace - List, select, and manage workspaces
• Task CRUD: anyt task create/get/update/delete - Full task lifecycle management
• Task Views: anyt task list/board - View tasks in lists or Kanban boards
• Dependencies: anyt task dep - Manage task dependencies
• Active Task: anyt task pick/drop - Track current task in progress
• AI Commands: anyt ai decompose - AI-powered task decomposition
• MCP Server: anyt mcp start - Run MCP server for Claude Code integration

Quick Start

Prerequisites

• Python 3.12+ with uv installed
• AnyTask Backend Server (API endpoint)
• Authentication: User token (JWT) or Agent API key

Installation

Option 1: Install from PyPI (Coming Soon)

pip install anyt

Option 2: Install from Source

git clone <repository-url>
cd AnyTaskCLI
make install

Initial Setup

1. Configure environment

   The CLI comes pre-configured with a production environment. To use a local development server instead:

   anyt env add dev http://localhost:8000
   anyt env use dev
   

   To verify your current environment:

   anyt env list
   

2. Authenticate

   For human users:

   anyt auth login
   # Paste your JWT token when prompted
   

   For AI agents:

   anyt auth login --agent-key anyt_agent_xxxxxxxxxxxxx
   

3. Select workspace

   anyt workspace list
   anyt workspace select DEV
   

4. Create your first task

   anyt task create "Implement user authentication" --status todo --priority 1
   

Usage Examples

View Tasks

# List all tasks
anyt task list

# Filter by status
anyt task list --status inprogress

# View as Kanban board
anyt task board

# Get task details
anyt task get DEV-123

Manage Tasks

# Create task
anyt task create "Add API endpoint" --status todo --priority 2

# Update task
anyt task update DEV-123 --status inprogress

# Add dependency
anyt task dep add DEV-124 --blocks DEV-123

# Delete task
anyt task delete DEV-123

Active Task Workflow

# Pick a task to work on
anyt task pick DEV-123

# View current active task
anyt task active

# Mark as done and drop
anyt task update DEV-123 --status done
anyt task drop

AI-Powered Commands

# Decompose a goal into tasks
anyt ai decompose "Implement user authentication system" --project-id 1

MCP Server for Claude Code

# Start MCP server
anyt mcp start

# Use with Claude Code by adding to claude_desktop_config.json:
# {
#   "mcpServers": {
#     "anytask": {
#       "command": "anyt",
#       "args": ["mcp", "start"]
#     }
#   }
# }

Claude Code Integration

AnyTask provides seamless integration with Claude Code through slash commands, enabling AI-assisted task management directly in your development environment.

Quick Setup

1. Install and configure AnyTask CLI (see Quick Start above)
2. Start using slash commands in Claude Code:
   • /anyt-next - Get intelligent task recommendations
   • /anyt-active - View your current task
   • /anyt-create - Create new tasks interactively
   • /anyt-board - See your Kanban board

Example Workflow

User: /anyt-next

Claude: "📋 Top Task Recommendations:

1. DEV-42: Implement OAuth callback (Score: 15.0)
   • Priority: 2 (Urgent)
   • All dependencies complete ✓
   • Unblocks 2 other tasks

Which task would you like to work on?"

User: "Let's do DEV-42"

Claude: "Great! Now working on DEV-42. Let me help you implement it..."

Features

• Smart Task Selection: AI analyzes priority, dependencies, and impact to recommend the best task
• Interactive Task Creation: Claude guides you through creating well-structured tasks
• Progress Tracking: View active tasks and board status without leaving your editor
• Context-Aware Help: Claude understands your current task and offers relevant assistance

Learn More

See Claude Code Integration Guide for:

• Complete setup instructions
• All available slash commands
• Workflow examples and best practices
• Troubleshooting tips

Two Integration Options:

1. CLI-based (recommended for getting started) - Simple slash commands, no extra setup
2. MCP Integration (advanced) - Full Model Context Protocol integration with real-time updates

Configuration

Environment Configuration

Configuration is stored in ~/.config/anyt/config.json:

{
    "environments": {
        "prod": {
            "api_url": "http://anyt.up.railway.app",
            "auth_token": "your-token-here",
            "default_workspace": 1
        },
        "dev": {
            "api_url": "http://localhost:8000",
            "auth_token": "your-dev-token-here",
            "default_workspace": 1
        }
    },
    "current_environment": "prod"
}

Workspace Configuration

Workspace settings are stored in .anyt/anyt.json:

{
    "workspace_id": "1",
    "name": "Development",
    "api_url": "http://localhost:8000",
    "last_sync": "2025-10-18T10:00:00Z"
}

Development

Setup Development Environment

# Clone repository
git clone <repository-url>
cd AnyTaskCLI

# Install dependencies
make install

# Run tests
make test

# Run type checking
make typecheck

# Format code
make format

# Lint code
make lint

Project Structure

src/
├── cli/                   # CLI application
│   ├── main.py           # Entry point
│   ├── client.py         # API client
│   ├── config.py         # Configuration management
│   └── commands/         # Command modules
│       ├── env.py
│       ├── auth.py
│       ├── workspace.py
│       ├── task.py
│       ├── ai.py
│       └── mcp.py
└── anytask_mcp/          # MCP server
    ├── server.py
    ├── tools.py
    └── resources.py

tests/
└── cli/
    ├── unit/             # Unit tests
    └── integration/      # Integration tests

Running Tests

# Run all tests
make test

# Run only unit tests
make test-cli-unit

# Run integration tests (requires server)
make test-cli-integration

Documentation

• CLI Usage Guide - Comprehensive CLI command reference
• Claude Code Integration - Using AnyTask with Claude Code slash commands
• MCP Integration - Advanced Model Context Protocol integration
• Development Guide - Developer guidelines and architecture

License

Proprietary - AnyTransformer Inc.

Support

For issues and feature requests, please contact: [e-mail]