A collection of agent skills for coding assistants like Claude Code and Codex, covering grilling sessions, TDD, bug diagnosis, planning and code review.
obra/superpowers
A skills framework and software development methodology that guides coding agents through brainstorming, planning, TDD, and reviewed subagent-driven implementation.
- Звёзды
- 301,846
- Форки
- 26,954
- Открытые issues
- 378
- Последний коммит
- 9 дней назад
Что это?
Superpowers is a set of composable skills and initial instructions that make a coding agent follow a structured software development workflow. When the agent detects that you are building something, it asks clarifying questions and presents a design in short sections for approval before writing any code.
After the design is approved, the agent writes an implementation plan broken into small tasks with exact file paths, code, and verification steps. Execution can dispatch a fresh subagent per task with review after each, or run tasks inline in one session with a final review. Skills such as test-driven development, systematic debugging, and code review trigger automatically during the work.
The project supports a wide range of coding agent harnesses, including Claude Code, Codex, Cursor, Gemini CLI, OpenCode, and Copilot CLI, each with its own installation path. It is distributed as a plugin or extension for each harness and is released under the MIT license.
Ключевые возможности
- 01Asks clarifying questions and presents the design in short chunks for sign-off before implementation.
- 02Produces implementation plans of small tasks with exact file paths and verification steps.
- 03Runs subagent-driven development with review after each task, or inline execution with a final review.
- 04Enforces red-green-refactor test-driven development during implementation.
- 05Includes skills for systematic debugging, code review, git worktrees, and branch completion.
- 06Installs as a plugin or extension for many coding agent harnesses.
Сценарии использования
- Having a coding agent build a feature autonomously while following a reviewed plan.
- Standardizing how a team's AI coding agents design, test, and review changes.
- Diagnosing why an agent skipped a skill or repeated work in a past session.
Темы
Частые вопросы
- Что такое superpowers?
- A skills framework and software development methodology that guides coding agents through brainstorming, planning, TDD, and reviewed subagent-driven implementation. Superpowers is a set of composable skills and initial instructions that make a coding agent follow a structured software development workflow. When the agent detects that you are building something, it asks clarifying questions and presents a design in short sections for approval before writing any code.
- superpowers можно использовать бесплатно? Какая у него лицензия?
- superpowers распространяется с открытым исходным кодом под лицензией MIT. Перед использованием в коммерческих продуктах проверьте условия лицензии.
- superpowers всё ещё поддерживается?
- Активно поддерживается. Последний коммит в superpowers — 10 окт. 2026 г..
- Для чего можно использовать superpowers?
- Having a coding agent build a feature autonomously while following a reviewed plan. Standardizing how a team's AI coding agents design, test, and review changes. Diagnosing why an agent skipped a skill or repeated work in a past session.
- Где скачать superpowers?
- Исходный код и инструкции по установке — на GitHub: https://github.com/obra/superpowers
Описание создано ИИ на основе README проекта. Статистика GitHub обновляется ежедневно.
Похожие проекты
Desktop app for browsing, installing and managing AI agent skills across 20+ coding agents from one visual interface.
A set of production-grade engineering workflow skills, packaged as Markdown, that guide AI coding agents through spec, plan, build, test, review, and ship.
A native macOS app for browsing, editing, and organizing AI coding agent skills and agents across Claude Code, Cursor, Codex, Windsurf, and Amp.
A modular orchestration system of plugins, agents and skills that automates task selection, code review, PR shipping and docs sync for AI coding agents.
A prompt-based skill that makes AI coding agents write only the code a task needs, favoring reuse and minimal solutions over large implementations.