architecture

Document Kubani architectural principles, patterns, and design decisions.

3|Updated Dec 2, 2025
One-click install
npx skills add https://github.com/X-McKay/kubani --skill architecture-x-mckay
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: architecture
Source: https://github.com/X-McKay/kubani/tree/main/.claude/skills/architecture
Command: npx skills add https://github.com/X-McKay/kubani --skill architecture-x-mckay

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a comprehensive understanding of Kubani's core architectural principles, design patterns, and component responsibilities, enabling developers to make informed decisions and contribute effectively to the project.

Core Features & Use Cases

  • Architectural Principles: Details key design philosophies like Agentic-First, Single Source of Truth, MCP-First Tool Integration, Registry-Centric Architecture, and Hierarchical Configuration.
  • Architectural Patterns: Explains patterns such as Federated Agent, Temporal Workflow, Memory Layer, and Learning Loop.
  • Component Responsibilities: Outlines the roles of core agents, specialized agents (like k8s-monitor, news-monitor), MCP servers, the Registry, and the UI.
  • Design Decisions: Justifies choices like using Temporal, MCP, Qdrant+Neo4j, and pydantic-settings.
  • Anti-Patterns: Highlights common pitfalls to avoid in development.
  • Use Case: A new developer joining the Kubani project can use this skill to quickly grasp the system's design and understand how different components interact, accelerating their onboarding process.

Quick Start

Understand the core architectural principles of Kubani by reviewing the architecture skill.

Frequently Asked Questions about architecture

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

FAQPage Schema
What are the core architectural principles for designing an agentic system?

Core architectural principles for agentic system design include Agentic-First, Single Source of Truth, MCP-First Tool Integration, Registry-Centric Architecture, and Hierarchical Configuration for robust component interactions.

How do federated agent and temporal workflow patterns work in system design?

Federated agent and temporal workflow patterns in system design coordinate distributed agents and manage long-running processes, while a memory layer and learning loop persist context and improve system responses over time.

How do I integrate MCP servers and Kubernetes monitoring agents into a registry-centric architecture?

Integrate MCP servers and k8s-monitor agents into a registry-centric architecture by registering specialized agents as distinct components, ensuring the Registry manages their lifecycle, tool integration, and hierarchical configuration.

Why choose Temporal, Qdrant, and Neo4j for agent system architecture?

Temporal, Qdrant, and Neo4j are chosen for agent system architecture to manage temporal workflows, provide vector search for memory layers, and structure complex component relationships within graph databases.

What anti-patterns should I avoid when building an MCP-first agent system?

Avoid anti-patterns in MCP-first agent systems by adhering to registry-centricity and hierarchical configuration, preventing decentralized tool integration, unmanaged agent lifecycles, and disconnected memory layers.

Does this architecture documentation cover component responsibilities for specialized agents like news-monitor?

The architecture documentation outlines component responsibilities for specialized agents like news-monitor and k8s-monitor, detailing their roles alongside core agents, MCP servers, the Registry, and the UI.