skills.rest
mguinada
mguinadaCOMMUNITY
Explore all skills by this author →
GitHub

ai-engineering

Design, implement, and debug agent patterns across MCP and ACI contexts.

Updated Feb 5, 2026
Search Tags:#routing#patterns#mcp#orchestration#workflows#agent-design#aci
One-click install
npx skills add https://github.com/mguinada/ai-coding-toolkit --skill ai-engineering-mguinada
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ai-engineering
Source: https://github.com/mguinada/ai-coding-toolkit/tree/main/skills/ai-engineering
Command: npx skills add https://github.com/mguinada/ai-coding-toolkit --skill ai-engineering-mguinada

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Complex AI systems require the ability to design, implement, and govern agentic workflows. This skill helps software teams decide when to use autonomous agents versus deterministic workflows, choose effective prompt patterns, and build tool-oriented architectures that are easy to test and maintain.

Core Features & Use Cases

  • Design and compare agentic patterns (routing, orchestrator-workers, memory management) for multi-tool workflows.
  • Integrate with MCP and ACI to enable plug-and-play tool ecosystems, with guardrails and observability.
  • Provide practical guidance for coding agents, data-analysis agents, and research agents, including testing, deployment, and monitoring.

Quick Start

Provide a high-level plan for building an agentic workflow that routes requests to specialized agents and applies chaining, routing, and memory patterns.

Frequently Asked Questions about ai-engineering

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

FAQPage Schema
How do I design agentic workflows with routing and orchestration patterns?▼

Design agentic workflows by implementing modular tool definitions and clear run-loop patterns that route requests to specialized agents. Teams can compare and build reliable agent patterns for multi-tool workflows, including routing, orchestrator-workers, and memory management.

When should I use autonomous agents versus deterministic workflows?▼

Choose autonomous agents versus deterministic workflows based on task complexity and reliability needs. This skill provides practical guidance to help software teams decide when autonomous agents are necessary and when deterministic workflows are more appropriate for maintainable architectures.

Does this skill support integrating tools with MCP and ACI contexts?▼

Yes, this skill supports integrating tools with MCP and ACI contexts. It enables plug-and-play tool ecosystems with guardrails and observability, allowing teams to build tool-oriented architectures that connect seamlessly across MCP and ACI environments.

What's the best way to test and debug agent patterns in multi-tool systems?▼

Test and debug agent patterns by applying evaluable testing practices and enforcing modular tool definitions. This skill helps teams implement clear run-loop patterns and memory management, ensuring agentic systems are easy to test, debug, and maintain across workflows.

Can I use this skill to build specialized coding and data-analysis agents?▼

Yes, you can build specialized coding and data-analysis agents. The skill provides practical guidance for coding agents, data-analysis agents, and research agents, covering testing, deployment, and monitoring within modular tool-oriented architectures.

Related Skills

Explore More Software Engineering Skills→
cornmanwtfcornmanwtf

ml-model-deployment

Deploy machine learning models with A/B testing, monitoring, and retraining.

Community
Advanced
vasilyu1983vasilyu1983
73

quality-debugging-troubleshooting

Diagnose software bugs and production incidents with systematic debugging workflows.

Community
Intermediate
0xUrsanomics0xUrsanomics
2

autoresearch

Diagnose execution traces and mutate prompts in a self-improvement loop.

Community
Advanced
penghang1223penghang1223
1

self-evolver

Automate self-monitoring, agent matching, task decomposition, and decision tracking.

Community
Advanced