agent-native-architecture

Official

Prompt-driven agents built from atomic primitives.

Author8b-is
Version1.0.0
Installs0

System Documentation

What problem does it solve?

It enables building agent-native architectures by centralizing behavior in prompts and atomic tools.

Core Features & Use Cases

  • Atomic tool primitives: read_file, write_file, list_files, and bash form the foundation for agent reasoning and orchestration.
  • Dynamic context injection: feed live app state into system prompts so agents act on current data.
  • Parity, composability and safety: ensure UI actions map to agent capabilities, allow gradual domain-tool graduation, and enable prompt-driven evolution or self-modification when appropriate.

Quick Start

  • Define a minimal set of primitives (read_file, write_file, list_files, bash) and craft a system prompt describing how the agent should use them.
  • Establish a shared workspace (Documents/ or equivalent) and a context.md file that the agent reads at session start.
  • Run the agent in a loop, observe tool calls, adjust the system prompt to refine judgment criteria and behavior.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: agent-native-architecture
Download link: https://github.com/8b-is/8b-is-mp/archive/main.zip#agent-native-architecture

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 223,000+ vetted skills library on demand.