skills.rest
teocns
teocnsCOMMUNITY
Explore all skills by this author →
𝕏@teocnsGitHub

skill-architecture

Design skill architectures with a three-level loading model for CLAUDE.md pointers.

Updated Mar 18, 2026
Search Tags:#monorepo#CLAUDE.md#references#skill-architecture#knowledge-routing#loading-model
One-click install
npx skills add https://github.com/teocns/proxify-cc-mirror --skill skill-architecture-teocns
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skill-architecture
Source: https://github.com/teocns/proxify-cc-mirror/tree/main/skills/skill-architecture
Command: npx skills add https://github.com/teocns/proxify-cc-mirror --skill skill-architecture-teocns

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Designs how to structure and wire CLAUDE.md and skills so knowledge actually reaches the agent at decision time, reducing wasted token budget and misrouting.

Core Features & Use Cases

  • Provides a three-level loading model (metadata, SKILL.md body, bundled resources) to balance context and depth.
  • Guides pointer patterns to route agents to the correct references and split responsibilities between CLAUDE.md and skills.
  • Promotes modular organization for scalable monorepos, enabling targeted, on-demand knowledge loading.

Quick Start

Identify the most relevant CLAUDE.md pointers and route agents to the right references to optimize instruction loading.

Frequently Asked Questions about skill-architecture

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I structure CLAUDE.md to ensure instructions reach the agent?▼

Structure CLAUDE.md using pointer patterns that route agents to correct references, splitting responsibilities between CLAUDE.md and skills to optimize instruction loading. This ensures knowledge reaches the agent at decision time.

What is the three-level loading model for agent skills?▼

The three-level loading model structures knowledge into metadata, SKILL.md body, and bundled resources. It balances context and depth to deliver guidance to agents while reducing wasted token budget and misrouting.

How do I organize skills for a monorepo with multiple domains?▼

Organize monorepo skills using modular organization and pointer-based routing discipline. This enables targeted, on-demand knowledge loading across multi-domain environments without overwhelming the agent context.

Why does my agent misroute or ignore CLAUDE.md instructions?▼

Agents misroute CLAUDE.md instructions when knowledge architecture lacks proper pointers and references. Without a structured loading model, context gets wasted and guidance fails to reach the agent at decision time.

When should I split responsibilities between CLAUDE.md and skills?▼

Split responsibilities when managing complex or multi-domain knowledge to prevent misrouting. Use CLAUDE.md for high-level pointers and skills for specific references, applying pointer-based routing discipline to optimize loading.

Related Skills

Explore More Software Engineering Skills→
HechaoYannetHechaoYannet

create-architecture

Generate a master architecture document from GDDs, ADRs, and governance constraints.

Community
Advanced
NinatuziNinatuzi

context-now

Provide real-time temporal context and deadline tracking to AI agents.

Community
Basic
msalganikmsalganik

skill-creator

Create or update Claude Skills with SKILL.md and YAML frontmatter.

Community
Intermediate
mikeparcewskimikeparcewski
8

context-engineering

Optimize context windows, memory scopes, and token usage in multi-agent systems.

Community
Advanced