co-reference

Loads the Cognitive Orchestration methodology reference covering eight principles, five layers, and six workflow phases.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams adopting AI-assisted workflows lack a shared, structured methodology for encoding institutional knowledge, guardrails, and human oversight, leading to inconsistent AI output and convention drift. ## Core Features & Use Cases - Methodology Reference: Distills the eight first principles, five-layer architecture, and six-phase workflow model of CO (Cognitive Orchestration) into a single self-contained reference. - Framework Mapping: Explains how CO connects to CARE (human role) and EATP (accountability), and how domain applications like COC, COR, and COG derive from the base methodology. - Use Case: When designing an AI-assisted development workflow, load this reference to understand which layer (Intent, Context, Guardrails, Instructions, Learning) should encode a given organizational concern, and how the six-phase Analyze-to-Deliver workflow structures work. ## Quick Start Ask the AI to explain the CO five-layer architecture and how the six-phase workflow maps to your project using the co-reference skill.

Frequently Asked Questions about co-reference

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

FAQPage Schema
What is Cognitive Orchestration (CO) methodology?▼

Cognitive Orchestration is a domain-agnostic methodology for structuring institutional knowledge, guardrails, and processes so AI agents produce trustworthy output under human oversight. It defines eight first principles, a five-layer architecture, and a six-phase workflow.

How do the five layers of CO architecture work?▼

The five layers are Intent (routing to specialized agents), Context (institutional knowledge), Guardrails (deterministic enforcement), Instructions (structured workflows with approval gates), and Learning (knowledge compounding across sessions). Each layer encodes a different aspect of human judgment.

How does CO relate to CARE and EATP frameworks?▼

CARE defines what the human is for, EATP defines how to keep the human accountable, and CO defines how the human structures AI work. CO inherits CARE's Human-on-the-Loop philosophy and maps its guardrails to EATP's Constraint Envelopes and Trust Postures.

What are the limitations of the CO methodology?▼

CO does not help in truly novel domains lacking institutional knowledge, does not solve the alignment problem, and depends on the quality of knowledge the human provides. Its three failure modes (amnesia, convention drift, safety blindness) reflect current AI limitations.

What is the six-phase workflow in CO?▼

The six phases are Analyze, Plan, Execute, Review, Learn, and Deliver, mapped to commands like /analyze, /plan, and /execute. Phase 05 (Learn) is unique because its output feeds back into the CO system itself, compounding institutional knowledge.