skills.rest
Victoriakaey
VictoriakaeyCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

agent-architecture

Guide engineers in selecting agent architectures and decision patterns.

Updated Apr 3, 2026
Search Tags:#workflow#architecture#agent#nodes#retrieval#observability#tool-use
One-click install
npx skills add https://github.com/Victoriakaey/build-reliable-agents --skill agent-architecture-victoriakaey
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-architecture
Source: https://github.com/Victoriakaey/build-reliable-agents/tree/main/skills/agent-architecture
Command: npx skills add https://github.com/Victoriakaey/build-reliable-agents --skill agent-architecture-victoriakaey

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Designing complex AI agent architectures often leads to poor observability, brittle pipelines, and hard-to-trace decision graphs. This Skill provides a structured framework to make architecture decisions early and document choices for teams.

Core Features & Use Cases

  • Decision: Workflow vs Agent for reliable retries and recovery
  • Decision: Single vs Multi-Agent for shared vs independent state
  • Decision: Tool-Use vs Specialized Nodes for clear traceability
  • Decision: Retrieval Strategy for data sources and safety layers
  • Use Case: Start a new LangChain-like agent project with clear architecture.

Quick Start

Create an architecture plan by explicitly deciding workflow vs agent, single vs multi-agent, tool-use vs nodes, and a retrieval strategy before coding.

Frequently Asked Questions about agent-architecture

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

FAQPage Schema
How do I design a multi-agent architecture with clear traceability and observability?▼

To design a multi-agent architecture with observability, you must explicitly decide on shared versus independent state management and impose guardrails for error handling. This ensures clear traceability across specialized nodes during production deployments.

What is the best way to choose between workflow vs agent patterns for reliable retries?▼

Choosing between workflow vs agent patterns requires evaluating your need for reliable retries and recovery. Workflows offer structured control for predictable paths, while agents provide dynamic decision-making for complex, variable task pipelines.

When should I use tool-use vs specialized nodes in an AI agent pipeline?▼

Use tool-use vs specialized nodes based on your traceability needs within the agent pipeline. Tool-use allows dynamic interaction with external functions, while specialized nodes provide fixed, clear traceability for specific isolated tasks.

How do I select a retrieval strategy for multi-agent pipelines with safety layers?▼

Selecting a retrieval strategy for multi-agent pipelines involves evaluating your data sources and implementing safety layers. This structured decision prevents brittle pipelines and ensures robust data retrieval across different agents.

Does this agent architecture framework support single-agent prototypes transitioning to production?▼

Yes, this architecture framework supports transitioning single-agent prototypes to production deployments. It guides engineers to impose explicit decisions on workflow patterns, tool-use, and retrieval strategies early to prevent brittle pipelines.

Related Skills

Explore More Software Engineering Skills→
melikhanmutlumelikhanmutlu

project-development

Map LLM-suited tasks into disciplined project architectures with prototype-driven evaluation.

Community
Advanced
Gami19Gami19

agent-ops

Automate post-deploy operations for AI agents with metrics, guardrails, and monitoring.

Community
Advanced
Owl-ListenerOwl-Listener
157

observability-design

Identify visibility requirements and capture execution traces for multi-agent workflows.

Community
Intermediate
markus41markus41
18

complex-reasoning

Apply Chain-of-Thought, Tree-of-Thought, and MECE analysis to complex problems.

Community
Advanced