superpowers

Superpowers

AI agents skip steps under time pressure. They bypass best practices when confident. They lack consistency across tasks.

Superpowers fixes this. Skills are mandatory instruction documents agents must follow. When a relevant skill exists, agents use it or fail the task.

How It Works

Skills are markdown files with proven workflows. The system teaches agents: "You have skills. Use them." This creates compliance, not optional reference.

Example: The test-driven-development skill forces RED-GREEN-REFACTOR. No test-first? Delete the code and start over. The skill prevents rationalization.

What's Inside

Testing

• test-driven-development, condition-based-waiting, testing-anti-patterns

Debugging

• systematic-debugging, root-cause-tracing, verification-before-completion, defense-in-depth

Collaboration

• brainstorming, writing-plans, executing-plans, dispatching-parallel-agents, requesting-code-review, receiving-code-review, using-git-worktrees, finishing-a-development-branch, subagent-driven-development

Meta

• writing-skills, sharing-skills, testing-skills-with-subagents, using-superpowers

Plus: Slash commands (/superpowers:brainstorm, /superpowers:write-plan, /superpowers:execute-plan) activate skills directly.

Philosophy

Test-driven development. Systematic over ad-hoc. Evidence over claims. Simplicity as primary goal.

Read more: Superpowers for Claude Code

Installation

Claude Code (via Plugin Marketplace)

In Claude Code, register the marketplace first:

/plugin marketplace add obra/superpowers-marketplace

Then install the plugin from this marketplace:

/plugin install superpowers@superpowers-marketplace

Verify Installation

Check that commands appear:

/help

# Should see:
# /superpowers:brainstorm - Interactive design refinement
# /superpowers:write-plan - Create implementation plan
# /superpowers:execute-plan - Execute plan in batches

Codex

Tell Codex:

Fetch and follow instructions from https://raw.githubusercontent.com/obra/superpowers/refs/heads/main/.codex/INSTALL.md

Detailed docs: docs/README.codex.md

OpenCode

Tell OpenCode:

Fetch and follow instructions from https://raw.githubusercontent.com/obra/superpowers/refs/heads/main/.opencode/INSTALL.md

Detailed docs: docs/README.opencode.md

Quick Start

Using Slash Commands

Brainstorm a design:

/superpowers:brainstorm

Create an implementation plan:

/superpowers:write-plan

Execute the plan:

/superpowers:execute-plan

Automatic Skill Activation

Skills activate automatically when relevant. For example:

• test-driven-development activates when implementing features
• systematic-debugging activates when debugging issues
• verification-before-completion activates before claiming work is done

What's Inside

Skills Library

Testing (skills/testing/)

• test-driven-development - RED-GREEN-REFACTOR cycle
• condition-based-waiting - Async test patterns
• testing-anti-patterns - Common pitfalls to avoid

Debugging (skills/debugging/)

• systematic-debugging - 4-phase root cause process
• root-cause-tracing - Find the real problem
• verification-before-completion - Ensure it's actually fixed
• defense-in-depth - Multiple validation layers

Collaboration (skills/collaboration/)

• brainstorming - Socratic design refinement
• writing-plans - Detailed implementation plans
• executing-plans - Batch execution with checkpoints
• dispatching-parallel-agents - Concurrent subagent workflows
• requesting-code-review - Pre-review checklist
• receiving-code-review - Responding to feedback
• using-git-worktrees - Parallel development branches
• finishing-a-development-branch - Merge/PR decision workflow
• subagent-driven-development - Fast iteration with quality gates

Meta (skills/meta/)

• writing-skills - Create new skills following best practices
• sharing-skills - Contribute skills back via branch and PR
• testing-skills-with-subagents - Validate skill quality
• using-superpowers - Introduction to the skills system

Commands

All commands are thin wrappers that activate the corresponding skill:

• brainstorm.md - Activates the brainstorming skill
• write-plan.md - Activates the writing-plans skill
• execute-plan.md - Activates the executing-plans skill

How It Works

1. SessionStart Hook - Loads the using-superpowers skill at session start
2. Skills System - Uses Claude Code's first-party skills system
3. Automatic Discovery - Claude finds and uses relevant skills for your task
4. Mandatory Workflows - When a skill exists for your task, using it becomes required

Philosophy

• Test-Driven Development - Write tests first, always
• Systematic over ad-hoc - Process over guessing
• Complexity reduction - Simplicity as primary goal
• Evidence over claims - Verify before declaring success
• Domain over implementation - Work at problem level, not solution level

Contributing

Skills live directly in this repository. To contribute:

1. Fork the repository
2. Create a branch for your skill
3. Follow the writing-skills skill for creating new skills
4. Use the testing-skills-with-subagents skill to validate quality
5. Submit a PR

See skills/meta/writing-skills/SKILL.md for the complete guide.

Updating

Skills update automatically when you update the plugin:

/plugin update superpowers

License

MIT License - see LICENSE file for details

Support

• Issues: https://github.com/obra/superpowers/issues
• Marketplace: https://github.com/obra/superpowers-marketplace