proxy-mode-reference

Name: proxy-mode-reference
Availability: InStock
Author: MadAppGang

Document claudish CLI invocation and routing for external AI models.

7|4|Updated Feb 26, 2026

Search Tags:#debugging #routing #cli #reference #claudish #external models

Ready to run

npx skills add https://github.com/MadAppGang/magus --skill proxy-mode-reference

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: proxy-mode-reference
Source: https://github.com/MadAppGang/magus/tree/main/plugins/multimodel/skills/proxy-mode-reference
Command: npx skills add https://github.com/MadAppGang/magus --skill proxy-mode-reference

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides essential reference information for understanding and utilizing external AI models through the claudish CLI, particularly within multi-model review workflows.

Core Features & Use Cases

External Model Invocation: Learn how to call external AI models directly via the claudish command-line interface.
Model Routing: Understand how claudish automatically routes requests to various providers like MiniMax, Kimi, GLM, Gemini, and OpenAI.
Debugging Integration: Troubleshoot issues related to external model integration and command execution.
Use Case: When /team command invokes an external model and you need to understand the underlying claudish command or debug why a specific model isn't being used correctly.

Quick Start

Execute the claudish --model grok-code-fast-1 --stdin --quiet < task.md > result.md command to invoke an external model.

Frequently Asked Questions about proxy-mode-reference

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

FAQPage Schema

How do I invoke external AI models using the claudish CLI?▼

To invoke external AI models using the claudish CLI, execute a command like `claudish --model grok-code-fast-1 --stdin --quiet < task.md > result.md` to route text prompts to your chosen provider.

How does model routing work for multi-model workflows in claudish?▼

Model routing in claudish automatically directs invocation requests to various external AI providers like MiniMax, Kimi, GLM, Gemini, and OpenAI based on deterministic patterns and specified model prefixes.

Why does the /team command invoke the wrong external model during routing?▼

If the /team command invokes the wrong external model, you can troubleshoot the underlying claudish command execution and review the deterministic invocation patterns to verify the correct model routing prefix is applied.

Can I use claudish for automated external model invocation via orchestrators?▼

Yes, you can use claudish for automated external model invocation via orchestrators like the /team command, which rely on the CLI's deterministic invocation patterns to execute multi-model review workflows.

What is the best way to debug claudish external model integration issues?▼

The best way to debug claudish external model integration issues is to reference the deterministic invocation patterns and routing prefixes to identify why a specific provider like Gemini or OpenAI is not being used correctly.