hive-concepts

Explains hive framework node types, graph edges, and Python package structure.

10.9k|5.7k|Updated Jan 12, 2026
One-click install
npx skills add https://github.com/adenhq/hive --skill hive-concepts
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hive-concepts
Source: https://github.com/adenhq/hive/tree/main/.claude/skills/hive-concepts
Command: npx skills add https://github.com/adenhq/hive --skill hive-concepts

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides foundational guidance for building goal-driven AI agents by detailing architecture, node types, tool discovery, and the overall workflow.

Core Features & Use Cases

  • Architectural Overview: Understand how Python packages are organized to host agent code, including exports, CLI, and modules.
  • Node & Edge Concepts: Learn about event_loop vs function nodes, and how edges drive graph execution with conditions and priorities.
  • Workflow Guidance: Learn best practices for client-facing nodes, node design, and graph validation to accelerate development.
  • Use Case: A new project team uses hive-concepts to bootstrap agent development and align on core principles before writing code.

Quick Start

Review this skill to grasp the hive architecture and then begin scaffolding a new agent with hive-create or by following the pattern in this guide.

Frequently Asked Questions about hive-concepts

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

FAQPage Schema
What are the foundational concepts for building goal-driven AI agents?

Foundational concepts for building goal-driven AI agents include architectural Python package organization, event_loop and function node definitions, tool discovery, and graph execution with conditions and priorities.

How do event_loop nodes differ from function nodes in graph execution?

Event_loop nodes and function nodes differ in their processing logic within the hive framework, where edges drive graph execution by routing between these nodes using conditions and priorities.

How do I structure a Python package for a new hive agent project?

Structure a Python package for a new hive agent by organizing modules, exports, and CLI components to host agent code, following architectural guidance for node design and graph validation.

What is the best way to design client-facing nodes in the hive framework?

The best way to design client-facing nodes in the hive framework is to follow workflow guidance and best practices for node design, ensuring proper graph construction and validation for architectural clarity.

Do I need prior hive framework knowledge to use this agent architecture guide?

You do not need prior hive framework knowledge to use this guide, as it is applicable to developers starting new projects or needing a refresher on node types, graph edges, and tool discovery.

How does tool discovery work across the hive framework?

Tool discovery across the hive framework works by codifying foundational concepts that allow goal-driven agents to locate and utilize necessary tools during graph execution and node processing.