llm-advisor

Consult multiple LLMs for second opinions on complex bugs and architecture decisions.

12|Updated Dec 13, 2025
One-click install
npx skills add https://github.com/ckorhonen/claude-skills --skill llm-advisor
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: llm-advisor
Source: https://github.com/ckorhonen/claude-skills/tree/main/skills/llm-advisor
Command: npx skills add https://github.com/ckorhonen/claude-skills --skill llm-advisor

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Enables teams to obtain second opinions from multiple LLMs (GPT-5.2, GPT-5, Gemini, Claude) to improve decisions on complex bugs, designs, and roadmaps.

Core Features & Use Cases

  • Multi-model consultation: Simultaneously consults several models to surface diverse perspectives on debugging, architecture choices, and planning trade-offs.
  • Structured comparison: Aggregates model recommendations, risks, and rationale to support defensible decisions.
  • On-demand escalation: Initiates cross-model reviews when single-model guidance is insufficient or biased.

Quick Start

Ask for a second opinion from multiple models on a complex issue.

Frequently Asked Questions about llm-advisor

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

FAQPage Schema
How do I get a second opinion from multiple LLMs on a complex architectural decision?

To get a second opinion from multiple LLMs on an architectural decision, you can use this Skill to simultaneously consult models like GPT and Gemini. It aggregates their recommendations, risks, and rationale to help you evaluate design trade-offs.

What is the best way to cross-reference LLM responses for debugging hard problems?

The best way to cross-reference LLM responses for debugging is to orchestrate prompts across multiple models simultaneously. This approach surfaces diverse perspectives and structured comparisons to resolve complex bugs faster.

Do I need the llm CLI to compare OpenAI and Gemini models?

Yes, you need the llm CLI and access to OpenAI and Gemini models to orchestrate prompts. This Skill requires the CLI to query multiple models and synthesize their guidance into a structured comparison.

When should I initiate a multi-model consultation for code review?

You should initiate a multi-model consultation for code review when single-model guidance is insufficient or biased, or when you are stuck on a hard debugging problem for more than 15 minutes.

Can I evaluate design trade-offs by querying multiple AI models at once?

Yes, you can evaluate design trade-offs by querying multiple AI models at once. This Skill consults several models to provide structured comparisons that support defensible decisions on complex designs and roadmaps.