Which models work best for coding tasks?

The skill recommends x-ai/grok-code-fast-1 for quick iterations and google/gemini-2.5-flash or openai/gpt-5 for complex reasoning and architectural planning.

Why does the skill insist on using sub-agents?

Running external models directly in the main context can add over 10,000 tokens of reasoning and transcript data, which quickly exhausts the context window. Sub-agents keep the main conversation lean.

What credentials do I need to use this skill?

You primarily need an OpenRouter API key. While an Anthropic API key is recommended to skip certain Claude Code dialogs, the model execution happens via OpenRouter.

What is the primary benefit of using Claudish with Claude Code?

Claudish allows you to use external models like Grok or Gemini within the Claude Code CLI interface, giving you access to diverse model strengths while keeping your project context managed.

Claudish Multi-Model Bridge

Name: Claudish Multi-Model Bridge
Author: involvex

byinvolvex

0•

Herramientas para Desarrolladores

Integrates OpenRouter models like Grok and GPT-5 into Claude Code workflows while preventing context window pollution.

Claudish-Usage is a specialized skill designed to extend Claude Code's capabilities by bridging it with the OpenRouter API. It enables developers to execute tasks using a wide array of external models, including Grok, Gemini, and GPT-5, directly within their terminal. The skill enforces a strict sub-agent delegation pattern to ensure that the main conversation context remains clean and efficient, preventing the token bloat typically associated with direct model proxying. With built-in agent selection logic and file-based instruction patterns, it provides a professional framework for multi-model AI development and complex code architectural tasks.

Características Principales

01File-based instruction patterns for complex, multi-phase coding workflows

02Direct agent invocation support via specialized CLI flags

03Strict sub-agent delegation to prevent context window token pollution

04Intelligent agent selection matrix for matching models to specific task types

05Multi-model support for Grok, Gemini, GPT-5, and MiniMax via OpenRouter

060 GitHub stars

Casos de Uso

01Performing cost-optimized code reviews using lightweight OpenRouter models

02Running rapid code iterations using Grok's high-speed reasoning

03Leveraging Gemini's massive context window for deep repository analysis

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add involvex/involvex-claude-marketplace claudish-usage

For use in Claude.ai and ChatGPT

Download Skill