Does this skill help with task validation?

Yes, it includes a Feedback Loop pattern designed to force Claude to verify results against success criteria before proceeding to the next step.

How does this help with AI agent performance?

By providing structured, machine-readable templates, it reduces ambiguity and ensures Claude follows specific operational patterns consistently.

What is the 'Exact Prompt' pattern?

It is a specialized structure for reproducible agent-to-agent handoffs, using encoded prompts to ensure tasks are passed between AI instances without loss of context.

Can I use this for security-critical workflows?

Yes, the Safety/Security archetype includes specific patterns for risk-tiering and explicit approval levels for destructive operations.

What are the four main archetypes provided by this skill?

The skill provides templates for CLI Reference (for tools), Methodology (for workflows), Safety/Security (for high-risk ops), and Orchestration (for multi-step processes).

Skill Archetypes

Name: Skill Archetypes
Author: majesticlabs-dev

bymajesticlabs-dev

•

Productivity & Workflow

Provides standardized structural templates and blueprints for building high-performance Claude Code skills.

Skill Archetypes serves as a meta-framework for developers looking to build custom Claude Code capabilities with professional-grade structure. It categorizes skill designs into four primary blueprints—CLI Reference, Methodology, Safety/Security, and Orchestration—ensuring that every new skill is optimized for AI comprehension and task execution. By following these proven patterns, developers can quickly implement robust features like tiered security approvals, agent-to-agent handoffs, and self-validating feedback loops without starting from scratch.

Key Features

01Pre-defined Markdown templates for consistent documentation and behavior

02THE EXACT PROMPT pattern for reliable agent-to-agent handoffs

03Four standardized skill blueprints for CLI, methodology, safety, and orchestration

04Integrated checklist and feedback loop patterns for process validation

05Risk-tiered approval structures for security-sensitive operations

0618 GitHub stars

Use Cases

01Creating safe DevOps automation skills with explicit human-in-the-loop triggers

02Designing a new custom Claude Code skill from scratch with a proven structure

03Converting internal engineering playbooks into executable AI methodologies

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add majesticlabs-dev/majestic-marketplace skill-archetypes

For use in Claude.ai and ChatGPT

Download Skill