Engineering
using-superpowers avatar

using-superpowers

The foundational skill for the Superpowers methodology. Ensures agents correctly identify and invoke required development skills before starting any task or conversation.

Introduction

The using-superpowers skill is the primary entry point for the Superpowers software development methodology. It is designed to act as a mandatory guardrail for coding agents, ensuring they do not bypass established workflows when handling user requests. By enforcing a strict policy of skill invocation before any substantive action—including clarifying questions or information gathering—it prevents agents from adopting ad-hoc or undisciplined approaches to complex programming tasks. This skill is intended for all software engineers and developers using automated coding assistants who want to guarantee consistent, high-quality output through rigorous adherence to predefined development processes.

  • Automatically triggers a check for relevant methodology skills upon receiving any user input.

  • Imposes a strict requirement to invoke specific tools like brainstorming, TDD (Test-Driven Development), or subagent-driven-development before proceeding.

  • Overrides default system behaviors that might encourage skipping workflows, ensuring project consistency.

  • Provides a clear decision framework to help agents determine when to pause, seek user clarification, or initiate a formal development sequence.

  • Integrates seamlessly with CLI environments including Claude Code, Gemini CLI, and GitHub Copilot CLI.

  • Always trigger this skill first when starting a new development session or addressing a new feature request.

  • Do not attempt to rationalize skipping the skill check; if there is a 1% chance it applies, execution is mandatory.

  • Follow the hierarchy of skill priority, placing process-oriented skills like brainstorming before implementation-specific skills like testing or deployment.

  • Use the provided interaction flow diagrams and red-flag checklists to self-audit decision-making during the task initiation phase.

  • Designed to function as the master controller, keeping the agent synchronized with the Superpowers ecosystem's best practices for red/green/refactor cycles and subagent orchestration.

Repository Stats

Stars
171,893
Forks
15,164
Open Issues
278
Language
Shell
Default Branch
main
Sync Status
Idle
Last Synced
Apr 29, 2026, 05:41 AM
View on GitHub