Can it improve my existing prompts?

The 'improve-prompt' mode analyzes your current skill configurations and execution logs to suggest optimizations based on real-world performance data.

How does the Orchestrate mode work?

Orchestrate mode allows the system to choose the best execution engine for a task, delegating to either Claude Code for deep codebase understanding or Codex for independent sub-tasks.

What is a Meta Skill Creator?

It is a specialized tool within Claude Code designed to build, manage, and refine other skills, ensuring they adhere to best practices like Domain-Driven Design and Continuous Delivery.

Does it manage software dependencies?

Yes, it creates 'Self-Contained Skills' that manage their own node_modules and package.json files using PNPM, preventing version conflicts with your main project.

Meta Skill Creator

Name: Meta Skill Creator
Author: ryryo

byryryo

•

Developer Tools

Automates the lifecycle of Claude Code skills through a collaborative, script-driven workflow for creating, updating, and optimizing AI capabilities.

About

Meta Skill Creator is an advanced utility designed to streamline the development of custom skills within the Claude Code environment. It employs a 'Collaborative First' philosophy, using interactive interviews to transform abstract ideas into concrete implementations while orchestrating tasks between different execution engines like Claude and Codex. By leveraging a library of 24 script types and a self-contained dependency management system using PNPM, it ensures that every generated skill is deterministic, portable, and capable of continuous self-improvement through integrated feedback loops.

Key Features

1 GitHub stars
Automated generation of self-contained skills with isolated dependencies
Progressive disclosure system to load resources only when needed for efficiency
Multi-engine orchestration between Claude Code and Codex for specialized tasks
Interactive collaborative mode for requirement gathering and skill design
Built-in feedback cycle that logs performance and suggests self-improvements

Use Cases

Refactoring existing AI prompts to improve deterministic outcomes and accuracy
Managing complex skill dependencies without polluting the main project environment
Rapidly prototyping new automation tools for specific team workflows

About

Key Features

1 GitHub stars
Automated generation of self-contained skills with isolated dependencies
Progressive disclosure system to load resources only when needed for efficiency
Multi-engine orchestration between Claude Code and Codex for specialized tasks
Interactive collaborative mode for requirement gathering and skill design
Built-in feedback cycle that logs performance and suggests self-improvements

Use Cases

Refactoring existing AI prompts to improve deterministic outcomes and accuracy
Managing complex skill dependencies without polluting the main project environment
Rapidly prototyping new automation tools for specific team workflows