AI/TLDR

Prime Radiant · 2026-06-16 · major

Superpowers v6.0 — Jesse Vincent's agentic-skills framework, ~2x faster reviews

Superpowers v6.0 reworks Jesse Vincent's open-source coding-agent skills framework so Claude Code and Codex finish reviews in about half the time and use roughly 50% fewer tokens, with one reviewer per task and three new harness integrations.

Superpowers GitHub repository social card by Jesse Vincent

v6.0 makes Superpowers' subagent-driven review loop run twice as fast and burn half the tokens.

Key specs

GitHub stars232k
Token savings vs v5~50%
Supported agents11+

Quick facts

MakerPrime Radiant (Jesse Vincent)
LicenseMIT
Stars232k
Latestv6.0.2 (June 17, 2026)
Install/plugin install superpowers@claude-plugins-official
Works withClaude Code, Codex, Cursor, Antigravity, Kimi Code, Pi, more

What is it?

Superpowers is a software-development methodology and skills framework for AI coding agents. It ships a fixed set of composable skills plus an instruction layer that pushes the agent through brainstorm, plan, test-driven implementation, and review. Version 6.0 is the latest major release of the framework, which has 232k GitHub stars.

How does it work?

The subagent-driven-development loop now dispatches one reviewer per task instead of two, with that reviewer returning both spec-compliance and quality verdicts at once. Diffs and task text move between controller and reviewer as files instead of inline pasted text, cutting context cost. The controller must explicitly pick a model for every dispatch, blocking silent upgrades to a more expensive tier. Reviewers run read-only and cannot be told to suppress findings.

Why does it matter?

Superpowers users running Claude Code or Codex through the framework now get the same review quality at roughly twice the speed and about 50% lower token spend, per Prime Radiant's release notes. The release also lands first-class harness support for three additional coding agents: Kimi Code (with marketplace install), Pi (via the native session-start extension), and Antigravity through its agy CLI with end-to-end test verification.

Who is it for?

Developers using Claude Code, Codex, Cursor, or any of the 11+ supported agents

Try it

claude /plugin install superpowers@claude-plugins-official

Sources

Tags

  • coding-agent
  • skills-framework
  • subagents
  • claude-code
  • codex
  • cursor
  • antigravity
  • kimi-code
  • open-source
  • mit-license

← All releases · Learn AI