Skip to content
← Tools / Skills
Skill · .md packageObra (community)/10 trustVerified

Superpowers

A complete agentic software-development methodology — brainstorming, TDD, subagent-driven execution — packaged as installable skills.

Visit official site →

About

Superpowers is a full development methodology for coding agents built on composable auto-triggering skills: it brainstorms the spec with the user, writes bite-sized implementation plans, enforces red/green TDD, and dispatches fresh subagents per task with code review between tasks.

The repo ships native plugin manifests for an unusually wide range of harnesses (Claude Code, Antigravity, Codex, Cursor, Gemini CLI, Copilot CLI, and more). MIT-licensed.

What it can do

  • Brainstorming skill (draws out the real spec before coding)
  • Writing-plans (2–5-minute tasks with exact file paths and verification steps)
  • Test-driven-development (strict RED-GREEN-REFACTOR enforcement)
  • Subagent-driven-development / executing-plans (per-task subagents with review)
  • Requesting-code-review and finishing-a-development-branch workflows

Setup

  1. Claude Code: /plugin install superpowers@claude-plugins-official
  2. Antigravity: agy plugin install https://github.com/obra/superpowers
  3. Cursor: /add-plugin superpowers · Codex CLI: /plugins → search "superpowers" · Gemini CLI: gemini extensions install https://github.com/obra/superpowers
  4. Reinstall per harness if you use several.

Prerequisites

One of the ~15 supported harnesses; no accounts or keys needed.

Use cases

codeautomationtesting

Works with

Claude CodeAntigravityCodexCursorCopilotAny agent (.md)

Trust score — 10/10

/10 trustVerified

The most-adopted skills repo in existence (~298K stars, 684 commits, 15 releases), MIT-licensed, maintained by a well-known open-source author, with per-harness install docs. One point off: it injects session-start hooks that reshape agent behavior (transparent in the repo, but heavier than a plain skill pack), and it's community rather than vendor-official.

★ 298,114 GitHub stars

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
10.0/10
Maintenance · recency, releases, issue load
10.0/10
Openness · license + docs
10.0/10
Manual review · sanity read + red-flag check
9.0/10
  • · 298,114 GitHub stars (log-scaled vs catalog peers)
  • · last push 3d ago; 15 releases/12mo; 334 open issues
  • · manual review: hooks reshape agent behavior — review before installing

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.