skills.rest
MadAppGang
MadAppGangOFFICIAL
Explore all skills by this author →
WebsiteGitHub

task-external-models

Invoke external AI models via Bash and the claudish CLI.

7|4|Updated Feb 26, 2026
Search Tags:#cli#llm#bash#orchestration#claudish#external model
One-click install
npx skills add https://github.com/MadAppGang/magus --skill task-external-models
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: task-external-models
Source: https://github.com/MadAppGang/magus/tree/main/plugins/multimodel/skills/task-external-models
Command: npx skills add https://github.com/MadAppGang/magus --skill task-external-models

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill simplifies the process of integrating and reliably invoking various external AI models (like Grok, Gemini, GPT-5) within orchestration workflows, ensuring deterministic execution via Bash and the claudish CLI.

Core Features & Use Cases

  • Reliable External Model Invocation: Use Bash commands with claudish for 100% reliable execution of external AI models.
  • Flag Passthrough: Seamlessly pass flags and parameters directly to underlying models like Claude Code.
  • Contextual Model Preferences: Automatically learn and reuse model preferences based on task context.
  • Use Case: When you need to compare the output of Grok and Gemini on a specific coding task, you can use this Skill to invoke both models via claudish and collect their results side-by-side.

Quick Start

Use claudish to run the grok-code-fast-1 model with the provided prompt.

Frequently Asked Questions about task-external-models

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

FAQPage Schema
How do I invoke external AI models reliably from a Bash orchestration script?▼

To invoke external AI models reliably from a Bash orchestration script, you can use the claudish CLI. It enables deterministic execution of diverse LLM providers within agentic systems.

Can I pass flags directly to Claude Code when using claudish for orchestration workflows?▼

Yes, you can pass flags directly to underlying models like Claude Code when using claudish. The Skill supports seamless flag passthrough to ensure parameters reach the target model.

What is the best way to compare outputs from multiple external LLMs in an agentic system?▼

The best way to compare outputs from multiple external LLMs in an agentic system is using claudish. It allows you to invoke models like Grok and Gemini, collecting their results side-by-side.

Does this approach support learning contextual model preferences for automated reuse?▼

Yes, the approach supports learning contextual model preferences for automated reuse. It automatically learns and applies these preferences based on the specific task context within your orchestration workflows.

Are there limitations when integrating Grok or Gemini into Bash workflows with claudish?▼

The metadata does not specify explicit limitations when integrating Grok or Gemini into Bash workflows with claudish. It is designed to provide 100% reliable execution for external model orchestration.

Related Skills

Explore More Software Engineering Skills→
MadAppGangMadAppGang
7

proxy-mode-reference

Document claudish CLI invocation and routing for external AI models.

Official
Basic
davpatel605-beepdavpatel605-beep

guidance

Constrain language model generation with regex and grammar rules for valid structured outputs.

Community
Advanced
PaulKinlanPaulKinlan
2

incident-classification

Classify, prioritize, and route incidents by severity and category.

Community
Advanced
attentiondotnetattentiondotnet

simpo-training

Optimize LLM preferences without a reference model using configurable training parameters.

Community
Advanced