skills.rest
RafayelGardishyan
RafayelGardishyanCOMMUNITY
Explore all skills by this author →
GitHub

agent-native-architecture

Build agent-native architectures with atomic tools and prompts in a loop.

Updated Mar 9, 2026
Search Tags:#workflow#architecture#prompts#tools#parity#granularity#agent-native
One-click install
npx skills add https://github.com/RafayelGardishyan/rafayels-marketplace --skill agent-native-architecture-rafayelgardishyan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-native-architecture
Source: https://github.com/RafayelGardishyan/rafayels-marketplace/tree/main/plugins/rafayels-engineering/.opencode/skills/agent-native-architecture
Command: npx skills add https://github.com/RafayelGardishyan/rafayels-marketplace --skill agent-native-architecture-rafayelgardishyan

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill enables building agent-native architectures where autonomous agents operate in a loop using atomic tools and prompts, eliminating the need to embed complex workflows directly in code.

Core Features & Use Cases

  • Parity: ensure every UI action has a corresponding agent tool, so users and agents can perform the same outcomes.
  • Granularity: break capabilities into atomic primitives the agent combines through prompts.
  • Composability: create new features by prompting, not by coding new workflows.
  • Emergent Capability: open-ended problems get tackled by tool-composition and reasoning within the loop.
  • Improvement Over Time: accumulate context and refine prompts for better behavior across sessions.

Quick Start

Define the atomic tools, craft a guiding system prompt, and run the agent loop until completion.

Frequently Asked Questions about agent-native-architecture

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

FAQPage Schema
What is an agent-native architecture and how does it work?▼

An agent-native architecture enables autonomous agents to operate in a loop using atomic tools and prompts, eliminating the need to embed complex workflows directly in code.

How do I ensure my application has parity between user and agent capabilities?▼

To achieve parity, ensure every UI action has a corresponding atomic agent tool, allowing both users and agents to perform the same outcomes within the application.

How do I build new features without coding complex workflows?▼

You can build new features through composability by breaking capabilities into atomic primitives and letting the agent combine them through prompts rather than coding new workflows.

Does this approach support accumulating context for agent improvement over time?▼

Yes, the architecture supports improvement over time by accumulating context and refining prompts, which leads to better agent behavior across multiple sessions.

What's the best way to start implementing an agent-native workflow?▼

To start, define atomic tools, craft a guiding system prompt, and run the agent loop until completion, optionally using sub-resources in scripts and assets directories.

Related Skills

Explore More Software Engineering Skills→
Vivekmano27Vivekmano27

agent-native-design

Map UI actions to atomic agent tools with parity mapping and runtime specifications.

Community
Advanced
fro-botfro-bot

agent-native-architecture

Model actions as atomic tools and outcomes as prompts for agent-native applications.

Community
Advanced
All-The-VibesAll-The-Vibes
50

agent-native-architecture

Design agent-native architectures using atomic tools and structured prompts.

Official
Advanced
ihj04982ihj04982

put-interaction-logic-in-event-handlers

Move React interaction logic directly into event handlers.

Community
Basic