lackey-mcp 3.1.0

Task chain management engine for AI agents with MCP integration

Lackey

AI agents that actually get things done. Task chain management with intelligent dependency handling

🎯 What is Lackey

Lackey turns AI agents from chatbots into project managers. Instead of endless back-and-forth, your AI agent breaks down complex goals into manageable task chains, tracks dependencies, and keeps everything organized—all stored directly in your repository.

30-second demo

# Install and initialize

python3 -m venv .venv && source .venv/bin/activate
pip install lackey-mcp && lackey init

# Launch Q chat with developer agent

q chat --agent developer

# Then in Q chat, tell the agent

# > "Build a REST API for user management with authentication"

# Watch it create a complete task chain

# ✅ Setup development environment

# ✅ Design database schema

# ⏳ Implement user model (depends on schema)

# ⏳ Add authentication (depends on user model)

# ⏳ Create API endpoints (depends on auth)

# ⏳ Write tests (depends on endpoints)

🚀 Quick Start

1. Install Lackey

# Create virtual environment

python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install Lackey

pip install lackey-mcp

2. Initialize Your Project

# Initialize project (creates Q chat agent configurations)

lackey init

# This creates

# - .lackey/ directory with project data

# - .amazonq/ directory with Q chat agent configurations

3. Start Working with AI Agents

# Launch Q chat with a specific agent role

q chat --agent manager

# Then interact within Q chat

# > "Let's plan our first project"

# > "Create a web development project for user management"

# Or use other agent roles

q chat --agent developer

# > "Get all ready tasks and start implementing authentication"

q chat --agent architect

# > "Design the database schema for our user system"

4. Watch the Magic

Your AI agent can now:

• Break down complex projects into task chains
• Manage dependencies automatically (no circular references!)
• Track progress and identify blockers
• Take notes and document decisions
• Keep everything in your git repository

✨ Why Lackey

Problem	Lackey Solution
🔄 Repetitive AI conversations	AI agents create persistent task chains
🧠 Agents forget context	All data stored in your repository
🔗 Complex dependencies	Automatic DAG validation prevents cycles
📊 No project visibility	Clear task status and progress tracking
🏢 Vendor lock-in	File-based storage, works with any AI agent

🛠️ Core Features

• 🤖 AI-First Design: Built specifically for AI agent workflows
• 📁 Repository Storage: All data lives in your project (no cloud required)
• 🔗 Smart Dependencies: Prevents circular dependencies with DAG validation
• 📝 Rich Notes: Comprehensive documentation with search
• 🎯 Zero Global State: Each project is completely self-contained
• 🔧 Extensible: Template system for different project types

📖 Documentation

New to Lackey? Start here:

• 📚 Getting Started Tutorial - Your first project in 5 minutes
• 🔧 AI Agent Integration - Connect different AI agents

Need specific help

• 🆘 Troubleshooting Guide - Common issues and solutions
• 📋 Task Patterns - Best practices and workflows

Looking up information

• 📖 API Reference - Complete API documentation
• 🛠️ CLI Commands - All command-line options
• 🔌 MCP Tools - All 25 tools + 3 gateways

Want to understand how it works

• 🏗️ Architecture Overview - System design principles
• 🚪 Gateway Pattern - Why we use gateways
• 🧠 Cognitive Load Management - AI optimization

📑 Full Documentation Index

🎬 Real-World Examples

Web Development Project

lackey init
q chat --agent manager

# > "Create an e-commerce website with user authentication, product catalog, shopping cart, and payment processing"

Data Science Pipeline

lackey init
q chat --agent manager

# > "Build a customer segmentation ML pipeline with data processing, model training, and deployment"

Content Creation

lackey init
q chat --agent writer

# > "Plan a 5-part technical blog series on microservices architecture with code examples"

🔌 MCP Integration

Lackey provides 3 semantic gateways that consolidate 25 individual tools:

• lackey_get - Retrieve information (projects, tasks, notes)
• lackey_do - Perform actions (create, update, manage)
• lackey_analyze - Analyze and optimize (dependencies, progress, bottlenecks)

This reduces AI agent cognitive load by 89% while maintaining full functionality.

🏗️ How It Works

your-project/
├── .lackey/                 # All Lackey data
│   ├── projects/           # Project definitions
│   ├── tasks/             # Task chains and metadata
│   ├── notes/             # Documentation and decisions
│   └── config.yaml        # Configuration
├── src/                   # Your actual code
├── docs/                  # Your documentation
└── README.md             # This file

Key principle: Your repository, your data. Delete .lackey/ anytime to remove all traces.

⚡ Advanced Features

Intelligent Dependencies

# Lackey prevents circular dependencies automatically

task_a.depends_on(task_b)
task_b.depends_on(task_c)
task_c.depends_on(task_a)  # ❌ Circular dependency detected and prevented

Rich Notes System

# AI agents can document their decisions

lackey.add_task_note(
    content="## API Design Decision\nUsing REST over GraphQL for simplicity",
    tags=["architecture", "api", "decision"]
)

Bulk Operations

# Efficiently manage multiple tasks

lackey.bulk_update_status(
    task_ids=[1, 2, 3, 4],
    status="in_progress",
    note="Starting sprint 2"
)

🚦 System Requirements

• Python: 3.10 or higher
• Storage: File-based (no database required)
• Platform: Windows, macOS, Linux
• AI Agent: Any MCP-compatible agent

🔧 Installation Options

Standard Installation

pip install lackey-mcp

Development Installation

git clone https://github.com/lackey-ai/lackey
cd lackey
pip install -e ".[dev]"
pre-commit install

With All Features

pip install lackey-mcp[docs,security]

🤝 Contributing

We welcome contributions! See our Development Guide for setup instructions.

# Quick development setup

git clone https://github.com/lackey-ai/lackey
cd lackey
pip install -e ".[dev]"
pytest  # Run tests

📄 License

Proprietary License - See LICENSE for details.

---

Ready to turn your AI agent into a project manager

pip install lackey-mcp && lackey serve

📚 Get Started → | 🆘 Get Help → | 🏗️ Learn More →