coc-reference

Explains the COC five-layer architecture for disciplined AI-assisted code generation.

Updated Apr 2, 2026
One-click install
npx skills add https://github.com/Hamza-Haadi/disease-risk-classifier-hamza --skill coc-reference-hamza-haadi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: coc-reference
Source: https://github.com/Hamza-Haadi/disease-risk-classifier-hamza/tree/main/.claude/skills/28-coc-reference
Command: npx skills add https://github.com/Hamza-Haadi/disease-risk-classifier-hamza --skill coc-reference-hamza-haadi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI coding assistants suffer from amnesia, convention drift, and security blindness because they lack institutional knowledge. This Skill provides the reference for the COC (Cognitive Orchestration for Codegen) framework, a five-layer architecture that replaces unstructured vibe coding with disciplined, context-aware AI development. ## Core Features & Use Cases - Five-Layer Architecture Reference: Covers Intent (30 specialized agents), Context (progressive disclosure skills), Guardrails (9 rules plus 9 hooks), Instructions (7-phase workflow with quality gates), and Learning (observation-instinct-evolution pipeline). - Vibe Coding Diagnosis: Explains the three fault lines of unstructured AI coding and why institutional knowledge matters more than model capability. - Autonomous Execution Model: Details the 10x throughput multiplier, structural versus execution gates, and Human-on-the-Loop developer responsibilities. - Use Case: When discussing AI-assisted development methodology or deciding how to structure agents, rules, and hooks in a Claude Code project, load this reference to ground the conversation in the COC framework. ## Quick Start Ask the AI to explain how the COC five-layer architecture prevents convention drift in AI-generated code.

Frequently Asked Questions about coc-reference

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

FAQPage Schema
What is the COC framework for AI-assisted coding?▼

COC (Cognitive Orchestration for Codegen) is a five-layer architecture for Claude Code that gives AI assistants institutional context, guardrails, and procedures. The layers are Intent, Context, Guardrails, Instructions, and Learning.

How does COC solve vibe coding problems?▼

COC addresses three fault lines: amnesia via persistent context engineering, convention drift via institutional knowledge replacing training defaults, and security blindness via deterministic hooks and rules that enforce compliance on every message.

What are the five layers of the COC architecture?▼

The layers are Intent (30 specialized agents), Context (28 skills with progressive disclosure), Guardrails (9 rules plus 9 enforcement hooks), Instructions (7-phase workflow with quality gates), and Learning (observation-instinct-evolution pipeline).

When should I use the coc-reference skill?▼

Load it when discussing AI-assisted development methodology, the five-layer architecture, anti-amnesia patterns, or Human-on-the-Loop development. For deeper analysis, invoke the coc-expert agent instead.

What are the limitations of the COC framework?▼

COC is less effective for novel architecture decisions, distributed systems complexity, legacy codebases without frameworks, and greenfield domains lacking institutional knowledge, where first sessions yield only 2-3x rather than 10x throughput.