Skill

using-superpowers

Use when starting any conversation — establishes skill usage rules and multi-agent routing

Install

Run in your terminal

npx claudepluginhub charliechan53/superpower-router --plugin superpower-router

Tool Access

This skill uses the workspace's default tool permissions.

Skill Content

Similar Skills

cache-components

Guides Next.js Cache Components and Partial Prerendering (PPR) with cacheComponents enabled. Implements 'use cache', cacheLife(), cacheTag(), revalidateTag(), static/dynamic optimization, and cache debugging.

cache-components

138.6k

claude-opus-4-5-migration

2 files

Migrates code, prompts, and API calls from Claude Sonnet 4.0/4.5 or Opus 4.1 to Opus 4.5, updating model strings on Anthropic, AWS, GCP, Azure platforms.

claude-opus-4-5-migration

83.2k

evaluation-methodology

1 file

Details PluginEval's skill quality evaluation: 3 layers (static, LLM judge), 10 dimensions, rubrics, formulas, anti-patterns, badges. Use to interpret scores, improve triggering, calibrate thresholds.

plugin-eval

32.9k

Stats

Stars0

Forks0

Last CommitMar 22, 2026

Actions

View Source View Plugin View on GitHub View README

Routing Default

For code construction and review tasks, route to Codex CLI (gpt-5.3-codex) via plan-and-execute.

For logic review, analysis, and research tasks, route to Codex CLI (gpt-5.4) via plan-and-execute.

For live web/docs research, use Claude's native WebSearch/WebFetch tools; stay on Claude.

For independent option gathering, run two Codex invocations in parallel (one with each model) via plan-and-execute.

Codex is fail-closed by default for code tasks: if Codex fails, ask the user for explicit approval before proceeding on Claude/Sonnet fallback.

Available Skills

Skill

When to use

brainstorming

Before any creative/feature work

test-driven-development

Before writing implementation code

systematic-debugging

When hitting bugs or unexpected behavior

plan-and-execute

Writing plans, executing tasks, routing to Codex (gpt-5.3-codex / gpt-5.4)

code-review

After implementation, or when receiving feedback

finishing-work

When claiming work is complete

writing-skills

Creating or editing skills

Using Skills

Rule: Check for relevant skills BEFORE any response or action. Use the Skill tool to invoke them.

Skill Priority

Process skills first (brainstorming, debugging) — determine HOW to approach
Implementation skills second (plan-and-execute, code-review) — guide execution

Routing Default

For code construction and review tasks, route to Codex CLI (gpt-5.3-codex) via plan-and-execute.
For logic review, analysis, and research tasks, route to Codex CLI (gpt-5.4) via plan-and-execute.
For live web/docs research, use Claude's native WebSearch/WebFetch tools; stay on Claude.
For independent option gathering, run two Codex invocations in parallel (one with each model) via plan-and-execute.
Codex is fail-closed by default for code tasks: if Codex fails, ask the user for explicit approval before proceeding on Claude/Sonnet fallback.

Available Skills

Skill	When to use
`brainstorming`	Before any creative/feature work
`test-driven-development`	Before writing implementation code
`systematic-debugging`	When hitting bugs or unexpected behavior
`plan-and-execute`	Writing plans, executing tasks, routing to Codex (`gpt-5.3-codex` / `gpt-5.4`)
`code-review`	After implementation, or when receiving feedback
`finishing-work`	When claiming work is complete
`writing-skills`	Creating or editing skills

Skill Types

Rigid (TDD, debugging): Follow exactly. Flexible (patterns): Adapt to context.