ai-systems-architect

What problem does it solve?

This Skill helps you architect reliable, production-ready AI/agent systems by preventing common failures like runaway tool loops, context overflow, weak evaluation, insecure tool access, and inefficient RAG vs fine-tuning choices.

Core Features & Use Cases

• End-to-end system architecture guidance: Aligns agent patterns, MCP server design, vector store selection, memory tiers, and evaluation strategy for building dependable applications.
• Practical decision frameworks: Chooses between routing vs reasoning models, RAG vs fine-tuning, and MCP transport options based on constraints like latency, update frequency, and deployment scale.
• Safety and cost guardrails: Defines anti-patterns and quality gates for iteration caps, timeouts, token budgeting, auth/rate limiting for tools, and golden-case evaluation.

Quick Start

Ask an AI to use ai-systems-architect to review your proposed agent design and produce an architecture with RAG/Routing decisions, MCP tool guardrails, context-window budgeting, and an evaluation plan.