claudekit / tools / superpowers
[ Plugin · Development ]

Superpowers

A complete software development methodology for your coding agents, built on composable skills that enforce TDD, systematic debugging, and plan-driven execution.

obra/superpowers ·published ·updated
$ /plugin install superpowers@claude-plugins-official copy

An open source project by Jesse Vincent (obra/superpowers, MIT).

What it does

Superpowers is a complete software development methodology for your coding agents, built on a set of composable skills plus initial instructions that make sure the agent actually uses them. It enforces true red/green TDD, YAGNI, and DRY, and lets Claude work autonomously for a couple of hours at a time without deviating from the plan.

Core skills

  • brainstorming — Refines rough ideas through questions, explores alternatives, presents design in sections for validation
  • writing-plans — Breaks work into bite-sized tasks (2-5 minutes each)
  • test-driven-development — Enforces RED-GREEN-REFACTOR
  • systematic-debugging — 4-phase root cause process
  • subagent-driven-development — Dispatches a fresh subagent per task with two-stage review (spec compliance, then code quality)
  • using-git-worktrees — Isolate feature work from current workspace
  • requesting-code-review / receiving-code-review — Structured review request and feedback
  • verification-before-completion — Forces evidence before success claims
  • dispatching-parallel-agents — Parallelizes 2+ independent tasks
  • writing-skills — Create new skills with progressive disclosure

Principles

  • Test-Driven Development
  • Systematic over ad-hoc
  • Complexity reduction
  • Evidence over claims

Typical flow

  1. brainstorming to refine the idea → writing-plans to break it down
  2. test-driven-development for the implementation loop
  3. systematic-debugging kicks in on any test failure or unexpected behavior
  4. verification-before-completion runs actual commands and confirms output before declaring done
§ 5

See also

same category · curated
[01]
[MCP] ARTEMIS · ARTEMIS turns natural-language instructions into reliable Android automation. Its native MCP server lets Claude Code drive real devices and emulators, run test workflows, and collect screenshots and Logcat output, while ARTEMIS's own agent performs the on-device steps in the background.
tool · claudekit.io / tools / artemis
[02]
[Tool] Claude Quickstarts · A collection of projects designed to help developers quickly get started with building deployable applications using the Claude API. Runnable quickstarts cover a customer support agent, a financial data analyst, computer use, browser automation, and Claude Managed Agents apps for Slack, Linear, Sentry, and MCP. Some quickstarts let Claude Code drive the setup.
tool · claudekit.io / tools / claude-quickstarts
[03]
[Tool] Worktrunk · A CLI for git worktree management, designed for running AI agents in parallel. Worktrees are addressed by branch name, and `wt switch -c -x claude feat` creates a worktree and starts Claude Code in it. The Claude Code plugin adds activity tracking in `wt list` and routes the worktrees agents create through Worktrunk.
tool · claudekit.io / tools / worktrunk
§ 6

Frequently Asked Questions

frequently asked
§ 6.1
What is Superpowers?
A plugin that auto-applies professional development workflows — brainstorming, TDD, systematic debugging, plan-driven execution, code review — as composable skills. An open source project by Jesse Vincent (`obra/superpowers`), listed on the claude-plugins-official marketplace.
§ 6.2
Where does it run?
Claude Code.
§ 6.3
How do I install it?
In Claude Code, run `/plugin install superpowers@claude-plugins-official`.
§ 6.4
Which skills are included?
brainstorming, test-driven-development, systematic-debugging, writing-plans, executing-plans, requesting-code-review, receiving-code-review, verification-before-completion, dispatching-parallel-agents, and more.
§ 6.5
Does it activate even on simple tasks?
Each skill triggers on conditions in its description. For example, verification-before-completion fires right before declaring a task complete, and systematic-debugging activates when a bug surfaces.