agent-native-architecture

Define agent outcomes in prompts using primitive tools.

Updated Dec 31, 2025

Search Tags:#prompts #prompt-native #agent-architecture #context-injection #primitives #dynamic-context

One-click install

npx skills add https://github.com/RBozydar/rbw-claude-code --skill agent-native-architecture-rbozydar

Or copy as Structured Prompt for Agent▼

Please help me install this Agent Skill.
Skill: agent-native-architecture
Source: https://github.com/RBozydar/rbw-claude-code/tree/main/plugins/core/skills/agent-native-architecture
Command: npx skills add https://github.com/RBozydar/rbw-claude-code --skill agent-native-architecture-rbozydar

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Traditional AI agents embed logic in code. This Skill promotes a prompt-native approach where outcomes are defined in prompts and the agent figures out how to achieve them using primitive tools. It enables flexible, maintainable agents that can evolve with user needs.

Core Features & Use Cases

Features are prompts; Tools are primitives
Dynamic capability discovery for external APIs
Context injection to reflect current app state
Self-modification principles for advanced agents
Clear lifecycle guidance for agent-based workflows

Quick Start

Define primitive tools (e.g., read_file, write_file, list_files) in the system prompt
Write a system prompt that describes outcomes and available primitives
Initialize and run the agent, feeding it runtime context