Superpowers is a complete software development methodology for your coding agents, built on top of a set of composable skills and some initial instructions that make sure your agent uses them.

Give your agent Superpowers with the universal prompt installer, or use the manual install for your specific AI coding tool.

It starts from the moment you fire up your coding agent. As soon as it sees that you're building something, it doesn't just jump into trying to write code. Instead, it steps back and asks you what you're really trying to do.

Once it's teased a spec out of the conversation, it shows it to you in chunks short enough to actually read and digest.

For long-running projects, it can maintain docs/superpowers/roadmap.md as a project-level map: product goal, current phase, milestones, backlog, risks, decisions, recent progress, and next recommended steps. Feature plans stay tactical; the roadmap keeps the project from drifting toward only the most recent context.

When work starts from PRDs, prototypes, meeting notes, user stories, or long product descriptions, it can turn that material into a development-ready requirement analysis for the current feature: scope, users, permissions, flows, business rules, acceptance criteria, engineering implications, and open questions.

After you've signed off on the design, your agent puts together an adaptive engineering plan. The plan fixes the goal, boundaries, risks, likely touchpoints, and verification strategy without pretending the final file layout or implementation code is knowable before inspecting the live code. It emphasizes YAGNI (You Aren't Gonna Need It), DRY, and risk-driven verification.

Next up, once you say "go", it launches a subagent-driven-development process, having agents work through each engineering task, inspect the live code before choosing final paths and abstractions, review their work, and replan when assumptions turn out to be wrong. It's not uncommon for Claude to be able to work autonomously for a couple hours at a time while staying aligned with the plan's intent.

There's a bunch more to it, but that's the core of the system. And because the skills trigger automatically, you don't need to do anything special. Your coding agent just has Superpowers.

If Superpowers has helped you do stuff that makes money and you are so inclined, I'd greatly appreciate it if you'd consider sponsoring my opensource work.

Thanks!

• Jesse

Installation differs by harness. If you use more than one, install Superpowers separately for each one.

Paste this prompt into your AI coding tool and let it install Superpowers Yeton for its own environment:

请帮我把 Superpowers Yeton 版安装到当前 AI 编程工具中。

仓库地址：
https://github.com/evepupil/superpowers-yeton-ver

要求：
1. 先识别我当前使用的 AI 编程工具，例如 Claude Code、Codex CLI、Codex App、Factory Droid、Gemini CLI、OpenCode、Cursor 或 GitHub Copilot CLI。
2. 根据当前工具选择正确安装方式：
   - Claude Code: /plugin install github:evepupil/superpowers-yeton-ver
   - Codex CLI: 打开 /plugins，添加 https://github.com/evepupil/superpowers-yeton-ver，然后安装插件
   - Codex App: 在 Plugins 里添加 https://github.com/evepupil/superpowers-yeton-ver 并安装
   - Factory Droid: 先运行 droid plugin marketplace add https://github.com/evepupil/superpowers-yeton-ver，再运行 droid plugin install superpowers@superpowers
   - Gemini CLI: gemini extensions install https://github.com/evepupil/superpowers-yeton-ver
   - OpenCode: 修改 opencode.json 的顶层 plugin 数组，添加 superpowers@git+https://github.com/evepupil/superpowers-yeton-ver.git
   - Cursor: /add-plugin https://github.com/evepupil/superpowers-yeton-ver
   - GitHub Copilot CLI: copilot plugin install github:evepupil/superpowers-yeton-ver
3. 如果已经安装官方 Superpowers，或者配置里存在 github.com/obra/superpowers，请先卸载、禁用或替换为 Yeton 版，避免同名 superpowers 插件冲突。
4. 保留我已有的其他插件、MCP、模型和项目配置，只修改 superpowers 相关配置。
5. 如果当前工具需要配置文件，请自动找到或创建对应配置文件，并只做最小修改。
6. 如果当前工具需要用户通过 UI 点确认，请告诉我要点哪里。
7. 安装完成后告诉我是否需要重启当前工具。
8. 重启后帮我验证是否能看到 Yeton 版新增 skill，例如 executing-roadmap、managing-project-roadmap、analyzing-requirements、creating-agents-guidance、qa-testing-workflow、qa-risk-review。
9. 如果你不确定当前工具的安装命令，请先读取这个仓库 README 或对应安��文档，再执行。

This fork keeps the plugin name superpowers, so install it as a replacement for the official package. Do not keep both the official obra/superpowers plugin and this fork enabled in the same harness.

Before installing this fork:

1. Remove or disable any existing official superpowers plugin in the current harness.
2. If the harness uses a config file, replace any obra/superpowers source with evepupil/superpowers-yeton-ver instead of adding a second superpowers entry.
3. Restart the harness after installation.
4. Verify the fork loaded by checking for fork-specific skills such as executing-roadmap, managing-project-roadmap, analyzing-requirements, creating-agents-guidance, qa-testing-workflow, or qa-risk-review.

Install this fork directly from GitHub:

/plugin install github:evepupil/superpowers-yeton-ver

Install this fork from its GitHub repository.

• Open the plugin search interface:

  /plugins

• Add the GitHub repository:

  https://github.com/evepupil/superpowers-yeton-ver

• Select Install Plugin.

Install this fork from its GitHub repository.

• In the Codex app, click on Plugins in the sidebar.
• Add or install from https://github.com/evepupil/superpowers-yeton-ver.
• Follow the prompts.

• Register the marketplace:

  droid plugin marketplace add https://github.com/evepupil/superpowers-yeton-ver

• Install the plugin:

  droid plugin install superpowers@superpowers

• Install the extension:

  gemini extensions install https://github.com/evepupil/superpowers-yeton-ver

• Update later:

  gemini extensions update superpowers

OpenCode uses its own opencode.json plugin config. Install Superpowers separately even if you already use it in another harness.

Edit your OpenCode config file:

• Windows: %USERPROFILE%\.config\opencode\opencode.json
• macOS/Linux: ~/.config/opencode/opencode.json

Add or update the top-level plugin array:

{
  "plugin": ["superpowers@git+https://github.com/evepupil/superpowers-yeton-ver.git"]
}

If an official Superpowers entry already exists, replace it with the fork URL above. Keep only one superpowers plugin entry enabled, then restart OpenCode.

• Detailed docs: docs/README.opencode.md

• In Cursor Agent chat, install this fork directly from GitHub:

  /add-plugin https://github.com/evepupil/superpowers-yeton-ver
  

Install this fork directly from GitHub:

copilot plugin install github:evepupil/superpowers-yeton-ver

1. executing-roadmap - Optional roadmap execution entry point. Reads docs/superpowers/roadmap.md, chooses one or more items, routes through planning, development, verification, roadmap update, and commit.

2. brainstorming - Activates before writing code. Refines rough ideas through questions, explores alternatives, presents design in sections for validation. Saves design document.

3. managing-project-roadmap - Maintains project-level direction, milestones, backlog, risks, progress, and next steps in docs/superpowers/roadmap.md.

4. analyzing-requirements - Activates when development starts from PRDs, prototypes, meeting notes, user stories, or long feature descriptions. Produces development-ready requirement analysis before planning.

5. using-git-worktrees - Activates after design approval. Creates isolated workspace on new branch, runs project setup, verifies clean test baseline.

6. writing-plans - Activates with approved design or requirement analysis. Links to the roadmap when present, then creates an adaptive engineering plan with goals, boundaries, known context, likely touchpoints, risks, and verification strategy.

7. subagent-driven-development or executing-plans - Activates with plan. Dispatches fresh subagent per task with two-stage review (spec compliance, then code quality), or executes in batches with human checkpoints.

8. test-driven-development - Activates during implementation when behavior changes need test-first development. It enforces RED-GREEN-REFACTOR for those changes; plans still choose verification by risk rather than forcing low-value mock tests everywhere.

9. requesting-code-review - Activates between tasks. Reviews against plan, reports issues by severity. Critical issues block progress.

10. verification-before-completion and finishing-a-development-branch - Verify completion, update roadmap when project-level progress changes, then present merge/PR/keep/discard options.

The agent checks for relevant skills before any task. Mandatory workflows, not suggestions.

Testing

• test-driven-development - RED-GREEN-REFACTOR cycle (includes testing anti-patterns reference)

QA Testing

• qa-testing-workflow - Staged QA workflow coordination with confirmation gates
• qa-requirement-analysis - Convert requirements and prototypes into testable QA materials
• qa-test-design - Business-oriented test point design
• qa-api-test-design - Optional API test design when code or API docs are available
• qa-risk-review - QA-focused review of code changes and test risk

Personal Context

• personal-info-cloudmind - Optional CloudMind-backed personal memory lookup for user preferences, history, project context, and prior decisions

Debugging

• systematic-debugging - 4-phase root cause process (includes root-cause-tracing, defense-in-depth, condition-based-waiting techniques)
• verification-before-completion - Ensure it's actually fixed

Collaboration

• executing-roadmap - Roadmap-driven execution entry point with guided and autopilot modes
• brainstorming - Socratic design refinement
• managing-project-roadmap - Project-level roadmap, progress, backlog, and next-step tracking
• analyzing-requirements - Development-ready requirement analysis before planning
• writing-plans - Adaptive engineering 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 two-stage review (spec compliance, then code quality)

Meta

• creating-agents-guidance - Create or complete project-level AI coding guidance files
• writing-skills - Create new skills following best practices (includes testing methodology)
• using-superpowers - Introduction to the skills system

• Risk-Driven Verification - Match tests, smoke checks, and manual probes to the real failure modes
• Systematic over ad-hoc - Process over guessing
• Complexity reduction - Simplicity as primary goal
• Evidence over claims - Verify before declaring success

Read the original release announcement.

The general contribution process for Superpowers is below. Keep in mind that we don't generally accept contributions of new skills and that any updates to skills must work across all of the coding agents we support.

1. Fork the repository
2. Switch to the 'dev' branch
3. Create a branch for your work
4. Follow the writing-skills skill for creating and testing new and modified skills
5. Submit a PR, being sure to fill in the pull request template.

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

Superpowers updates are somewhat coding-agent dependent, but are often automatic.

MIT License - see LICENSE file for details

Superpowers is built by Jesse Vincent and the rest of the folks at Prime Radiant.

• Discord: Join us for community support, questions, and sharing what you're building with Superpowers
• Issues: https://github.com/evepupil/superpowers-yeton-ver/issues
• Release announcements: Sign up to get notified about new versions