How can I run multiple models at the same time?

You can run models in parallel using the Bash CLI with the '&' operator or by initiating multiple background tasks using the PROXY_MODE prompt directive.

How do I find the correct model ID for an external LLM?

Use the command 'claudish --top-models' to see current paid models or 'claudish --free' to see available free models and their IDs.

Which agents support the PROXY_MODE prompt directive?

Specific agents in the agentdev, frontend, seo, and dev plugins support PROXY_MODE. If an agent is not supported, you should use the Bash CLI approach instead.

Is PROXY_MODE a parameter for the Task tool?

No, PROXY_MODE is not a Task parameter. It must be placed on the first line of the prompt string itself to be recognized by the agent.

External Model Orchestrator

Name: External Model Orchestrator
Author: MadAppGang

byMadAppGang

•

211

Productivity & Workflow

Enables the integration and management of external AI models like Grok, Gemini, and GPT-5 within Claude Code tasks.

About

This skill provides a vital quick-reference guide for developers looking to expand Claude's capabilities with external Large Language Models (LLMs). It clarifies the essential distinction between tool parameters and prompt directives, specifically teaching users how to implement PROXY_MODE correctly within supported agents or utilize the universal Bash CLI for any agent. By providing standardized patterns for parallel execution and model-specific routing, it empowers users to orchestrate multi-model workflows for comprehensive code reviews, debugging, and architectural analysis.

Key Features

Correct PROXY_MODE syntax guidance for prompt-based model switching
Compatibility matrix for agents supporting PROXY_MODE directives
211 GitHub stars
Universal Bash CLI patterns for running any agent with external LLMs
Quick commands for identifying current available free and paid models
Parallel multi-model execution workflows for simultaneous reviews

Use Cases

Automating multi-model validation workflows within background tasks
Leveraging specialized coding models like Qwen or DeepSeek for niche language debugging
Performing parallel code reviews across Grok, Gemini, and GPT-5 for diverse architectural feedback

About

Key Features

Correct PROXY_MODE syntax guidance for prompt-based model switching
Compatibility matrix for agents supporting PROXY_MODE directives
211 GitHub stars
Universal Bash CLI patterns for running any agent with external LLMs
Quick commands for identifying current available free and paid models
Parallel multi-model execution workflows for simultaneous reviews

Use Cases

Automating multi-model validation workflows within background tasks
Leveraging specialized coding models like Qwen or DeepSeek for niche language debugging
Performing parallel code reviews across Grok, Gemini, and GPT-5 for diverse architectural feedback