skills.rest
ahostbr
ahostbrCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

ls-consult

Query configured external model endpoints and CLI models to synthesize responses for a single question.

Updated Aug 5, 2026
Search Tags:#multi-model#claude-code#architecture-review#external-apis#prompt-synthesis#llm-consultation#cross-model-comparison
One-click install
npx skills add https://github.com/ahostbr/liteharness-plugin --skill ls-consult
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ls-consult
Source: https://github.com/ahostbr/liteharness-plugin/tree/main/skills/ls-consult
Command: npx skills add https://github.com/ahostbr/liteharness-plugin --skill ls-consult

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you get a second opinion or comparative perspective on a decision, design, or reasoning by consulting multiple external LLMs and synthesizing their outputs.

Core Features & Use Cases

  • Multi-model consultation: Sends your question to multiple configured models (including both HTTP/API and CLI-driven models) and returns attributed results.
  • Config-driven model selection: Supports presets and explicit model overrides via --preset and --models, plus shared system prompt controls.
  • Live CLI streaming + polling: For CLI-backed models, launches visible windows, captures streamed output, and polls until results are ready or timed out.
  • Use-case fit: Ideal when the user explicitly asks to “consult” specific external models (e.g., GPT/Gemini) or requests cross-model comparison; it is not meant for general internal code review (use /consult-polymaths instead).

Quick Start

Ask the skill to compare opinions by running /consult with your question, for example: “/consult Can you consult GPT and Gemini on whether this database schema makes sense?”.

Frequently Asked Questions about ls-consult

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

FAQPage Schema
How do I get multiple LLM opinions on a database schema design?▼

Multi-model consultation resolves schema design questions by querying configured external LLM endpoints and CLI models in parallel, then synthesizing their responses to provide comparative technical perspectives for decision-making.

What is the best way to compare external LLM responses for architecture tradeoffs?▼

Comparing external LLM responses for architecture tradeoffs is best handled by a consultation skill that sends your prompt to configured model providers, performs HTTP health checks, and returns attributed results from each model.

How do I configure external model endpoints for cross-model comparison?▼

You configure external model endpoints for cross-model comparison by defining settings in a consult-config.json file, which allows you to specify model providers, apply presets, and manage shared system prompts for your queries.

Can I use CLI-driven models for LLM consultation alongside HTTP APIs?▼

Yes, you can use CLI-driven models alongside HTTP APIs for LLM consultation; the system launches visible CLI processes, captures streamed output, and polls prompt or output files until results are ready or timed out.

When should I not use a cross-model comparison tool?▼

You should not use a cross-model comparison tool for general internal code reviews, as these consultation features are specifically designed for scenarios where you explicitly request external model opinions or architecture tradeoff analysis.

Related Skills

Explore More Software Engineering Skills→
databricksdatabricks
13

databricks-model-serving

Manage Databricks Model Serving endpoints via CLI for LLMs and custom models.

Official
Advanced
haryelramalhoharyelramalho

council

Orchestrate multi-advisor debates to produce synthesized recommendations with risks.

Community
Advanced
bettercallfanbettercallfan

opengrep-compliance

Automate code compliance scanning with OpenGrep rules and LLM analysis.

Community
Advanced
AIsa-teamAIsa-team
23

llm-router

Route chat, vision, and comparison requests across 70+ AI models via one API key.

Official
Basic