• Pricing
  • Latest
Sign InGet started free
Home
Product
Studio overview

Plan the work

  • Goals & Initiatives
  • Research Agents
  • Briefs
  • Plans

Deliver together

  • Cloud Agents
  • Hamster CLI
  • Collaboration
  • Issue tracker sync

Share knowledge

  • Context Graph
  • Blueprints
  • Skills & Methods
  • Routines
  • Connections
PricingLatest
DocsResearchCareersAboutThe Hamster Method
Sign InGet Started

Turn your team's goals into delivered work.

© 2026 Wheel Go Fast, Inc. All Rights Reserved.

GitHubEmail support
Product
  • Studio overview
  • Goals & Initiatives
  • Research Agents
  • Briefs
  • Plans
  • Cloud Agents
  • Hamster CLI
  • Collaboration
  • Issue tracker sync
  • Context Graph
  • Blueprints
  • Skills & Methods
  • Routines
  • Connections
  • Pricing
  • Method
For
  • Developers
  • Founders
  • Product managers
  • Designers
  • Agents
Works with
  • Claude Code
  • Cursor
  • Codex
  • Copilot
  • Gemini CLI
  • Grok
  • Grok Bot
Resources
  • Research
  • Methods
  • Skills
  • Docs
  • Latest
  • Changelog
  • FAQ
  • About
  • Careers
Legal
  • Terms of Service
  • Privacy Policy
  • Cookie Policy
  • Trust Center

© 2026 Wheel Go Fast, Inc. All Rights Reserved.

GitHubEmail support
  • Pricing
  • Latest

Research/Taskmaster

Structure for AI. Sanity for you.

AI agents are fast but directionless. Taskmaster gives them a structured task list with dependencies, context, and acceptance criteria — so they build what you actually need.

Try TaskmasterTry Hamster
curl -fsSL https://tryhamster.com/cli/install | bash
zsh — taskmaster
→~/hamstertm list
IDPRIORITYTITLESTATUS
ABC-120lowSetup repo structure✓ done
ABC-123highImplement OAuth Flow○ pending
└— 123.1-Setup Google Strategy○ pending
ABC-124mediumUser Dashboard UI! blocked

Stop building the wrong thing fast.

The tools are powerful, but the workflow is broken.

Linear
Figma
Slack
GitHub
⚠️ Context Switching...

Context Hunting

You spend more time jumping between Linear tickets, Figma designs, and Slack threads than actually coding.

prompt: "Build auth system"
Auth0?
JWT?
Firebase?
AI deciding implementation details...

AI Decides How

A simple prompt leaves the implementation details to the AI. Without a plan, it builds the wrong thing, fast.

Intent
tm next
Shipped

No Standard Flow

Vibe coding is fun for side projects. But teams need a predictable path from intent to shipped code.

zsh — taskmaster
requirements.txt (PRD)
Build a user authentication system with email/password and OAuth...

PRD → Tasks → Subtasks → Done

Start with requirements. End with shipped code. Taskmaster manages everything in between.

tm parse-prd

Turn any doc into structured tasks

tm expand

Break complex tasks into subtasks

tm next

Get the next task with full context

tm set-status

Track progress as you ship

Know what needs to be broken down.

tm analyze-complexity scores each task 1–10 and recommends which ones to expand. Stop guessing. Let AI tell you where the risk is.

Explore Complexity Analysis→
zsh — taskmaster
→tm analyze-complexity
Analyzing 5 tasks...
1.
Setup repo structure
2
2.
Implement OAuth
7
3.
Build payment system
9
4.
Add user dashboard
5
5.
Write unit tests
3
⚠️ Recommend expansion: Tasks 2, 3 (complexity ≥ 7)

Fresh info beyond the cutoff.

AI models have knowledge cutoffs. tm research queries live sources for the latest best practices, security patches, and library updates.

Explore Research→
zsh — taskmaster
→tm research "React Query v5 best practices"

Test-driven development on autopilot.

A state machine that runs the full Red → Green → Refactor cycle. Creates branches, runs tests, retries on failure. You review. It ships.

zsh — taskmaster
→~/repotm autopilot start ABC-123
[autopilot] Initializing TDD workflow for task ABC-123...
[autopilot] Created branch task/abc-123-oauth

Local or connected. Your choice.

Taskmaster works standalone with your own API keys. Connect to Hamster for team sync and hosted AI.

Local Mode

Fully offline, your keys

  • Parse PRDs locally with tm parse-prd
  • Tasks stored in .taskmaster/ directory
  • Bring your own API key (Anthropic, OpenAI, Gemini, Ollama, etc.)
  • Works with Claude Code & Gemini CLI (no key needed)
Configure via tm models --setup

Connected to Hamster

Team sync, hosted AI

  • Briefs become tasks via tm briefs
  • Sync local work to cloud with tm export
  • No API keys needed — inference via Hamster
  • Invite teammates to collaborate on briefs
Your PM plans in Hamster, you execute in terminal

Briefs become tasks in seconds.

Pull down team plans instantly. No context switching. No copy-pasting requirements from Notion.

zsh — taskmaster
→~/projecttm briefs

Isolated contexts for branches.

Working on multiple features? Use tags to keep task lists separate. No merge conflicts. No crossed wires. Switch contexts instantly.

NEXT TASK
ID: ABC-123

Implement Auth

Set up the authentication flow using Supabase Auth. Include email/password and Google provider.

Priority
High
Complexity
5/10
Suggested Actions
→tm set-status -s in-progress
→tm show --json | cursor

Works where you work.

CLI-first means it works everywhere. Native MCP support for AI IDEs.

Explore MCP Server→
Cursor
MCP
VS Code
MCP
Windsurf
MCP
Claude Code
Native
Any Terminal
CLI

Open source. Always.

Taskmaster is MIT licensed. Read the code, contribute, fork it. The core will always be free.

Star on GitHubView Changelog
28k+
GitHub Stars
2M+
npm Downloads

Everything else you need.

PRD Parser

Transform requirements docs into structured, executable tasks with dependencies and priorities.

Complexity Analysis

AI scores each task 1–10 for complexity and recommends which ones to break down before you start.

Task Expansion

Automatically expand high-level tasks into detailed subtasks with file references and acceptance criteria.

Research Tool

Query live sources for the latest best practices, security patches, and library updates.

MCP Server

Expose every CLI command as an MCP tool your AI coding assistant can call directly.

Bring Your Own Key

15+ AI providers, zero-key CLI tools, local models, or zero-config when connected to Hamster.

Taskmaster proved planning helps. Studio is where your team aligns.

Taskmaster gives individual developers structured task management for AI agents. Studio brings the whole team into the picture — collaborative briefs, real-time alignment, and a shared plan everyone executes against.

Already using Taskmaster? Studio is where your team aligns before agents execute. The CLI connects both.

Explore Hamster Studio→

Taskmaster

Solo developer + AI agent

Parse PRDs, manage tasks, execute with your coding agent. Works offline with your own API keys.

↓

Hamster Studio

Team alignment + AI execution

Collaborative briefs, real-time editing, team plans, issue tracker sync, and the same CLI your agents already use.

Give your AI agents structure.

Install Taskmaster in seconds. Start shipping code that actually matches the requirements.

Try TaskmasterTry Hamster
Sign In
Get Started
Try Hamster
Try Hamster

Turn your team's goals into delivered work.

© 2026 Wheel Go Fast, Inc. All Rights Reserved.

GitHubEmail support
Product
  • Studio overview
  • Goals & Initiatives
  • Research Agents
  • Briefs
  • Plans
  • Cloud Agents
  • Hamster CLI
  • Collaboration
  • Issue tracker sync
  • Context Graph
  • Blueprints
  • Skills & Methods
  • Routines
  • Connections
  • Pricing
  • Method
For
  • Developers
  • Founders
  • Product managers
  • Designers
  • Agents
Works with
  • Claude Code
  • Cursor
  • Codex
  • Copilot
  • Gemini CLI
  • Grok
  • Grok Bot
Resources
  • Research
  • Methods
  • Skills
  • Docs
  • Latest
  • Changelog
  • FAQ
  • About
  • Careers
Legal
  • Terms of Service
  • Privacy Policy
  • Cookie Policy
  • Trust Center

© 2026 Wheel Go Fast, Inc. All Rights Reserved.

GitHubEmail support
Sign InGet started free
Sign In
Get started free