How does CSO improve my custom skills?

CSO (Search Optimization for LLMs) helps you write descriptions and triggers that act like SEO for AI models, making it much more likely that the agent correctly identifies and loads the right skill for a given task.

What are the different architecture tiers?

Tier 1 is for simple single-file skills (<200 lines), Tier 2 is for complex multi-concept skills (up to 1000 lines), and Tier 3 is for massive platform-level integrations.

What is 'Anti-Rationalization' in AI skills?

Anti-Rationalization refers to a set of writing techniques that prevent an AI agent from making excuses or 'rationalizing' why it didn't follow a specific rule, ensuring strict adherence to discipline-based instructions.

Is this skill compatible with Cursor?

Yes, while designed with Claude Code in mind, these standards and templates are battle-tested for Antigravity, Cursor, and other agentic AI coding environments.

Skill Architecture & Writing

Name: Skill Architecture & Writing
Author: sickn33

bysickn33

•

15,684

•

Productivity & Workflow

Standardizes the creation, optimization, and documentation of high-performance AI agent skills using tiered architectures and anti-rationalization patterns.

This meta-skill serves as a comprehensive framework for developers looking to build, debug, and scale custom capabilities for Claude Code and other agentic environments. It provides a structured decision tree to help users choose between simple, expanded, or platform-level architectures, ensuring skills remain modular and maintainable. By incorporating advanced techniques like 'Anti-Rationalization' to prevent instruction drift and 'CSO' for improved model triggering, this skill ensures that custom agent capabilities are both discoverable and reliable during complex development tasks.

Key Features

01Standardized testing guides and RED-GREEN-REFACTOR cycles for skills

02Pre-deploy checklists for metadata, naming, and structural compliance

03Tiered architecture templates for simple to platform-level skills

04Anti-Rationalization patterns to improve AI instruction adherence

05CSO (Search Optimization for LLMs) to ensure accurate skill triggering

0615,684 GitHub stars

Use Cases

01Debugging and fixing instances where AI agents ignore specific skill rules

02Standardizing skill documentation and metadata across a large organization

03Developing custom team-specific agent capabilities from scratch

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add sickn33/antigravity-awesome-skills writing-skills

For use in Claude.ai and ChatGPT

Download Skill