What kind of output does the diagnostic tool provide?

It provides the agent name, the operational status, the response latency, and the raw text response or error message from the model.

Which AI models are supported by the agent skill?

The skill supports Claude, Codex, Copilot, Cursor, and Gemini models, provided the corresponding MCP tools are configured.

Does this skill modify my source code?

No, the Quorum Agent skill explicitly prepends a read-only instruction to all prompts, ensuring agents only analyze and never modify your files.

How do I test all available agents at once?

You can use the command '/quorum:agent all ' to trigger all enabled agents in parallel and see their raw outputs.

Do I need separate API keys for each agent?

Yes, this skill relies on external MCP tools (like quorum-copilot or quorum-gemini) which must be individually configured with their respective API credentials.

Quorum Multi-Model Agent Runner

Name: Quorum Multi-Model Agent Runner
Author: nexusentis

bynexusentis

•

数据科学与机器学习

Executes diagnostic queries across multiple AI agents including Claude, Copilot, and Gemini to compare raw model outputs within your codebase.

The Quorum Agent skill acts as a powerful diagnostic bridge within the Quorum framework, allowing users to trigger specific AI models like Claude, Codex, Copilot, Cursor, or Gemini directly from the command line. It operates in a mandatory read-only analysis mode, making it an essential tool for testing model responses to technical prompts and exploring codebases without the risk of accidental modifications. By delivering raw, unreconciled output alongside latency data, it provides developers with a transparent way to benchmark model performance and accuracy across different LLM providers.

主要功能

011 GitHub stars

02Parallel execution capability to compare all enabled agents simultaneously

03Strict read-only execution mode for safe codebase exploration

04Support for Claude, Codex, Copilot, Cursor, and Gemini models

05Direct MCP tool integration for seamless external agent querying

06Real-time status and latency reporting for every agent response

使用场景

01Performing non-destructive architectural reviews and codebase diagnostics

02Testing prompt engineering effectiveness across multiple LLMs at once

03Comparing code analysis capabilities between GitHub Copilot and Google Gemini on specific files

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add nexusentis/quorum agent

For use in Claude.ai and ChatGPT