How are API errors and rate limits handled?

The skill includes built-in error handling that automatically retries transient issues such as 429 rate limits or 502/503 server errors.

Why does this skill require disabling the sandbox on macOS?

On macOS, the UV tool requires access to system configuration frameworks to detect network proxy settings, which can trigger sandbox restrictions.

Can I generate images with transparent backgrounds?

Yes, by using compatible models like gpt-5-image and specifying the --background transparent and --output-format png flags.

Does this skill support model discovery?

Yes, you can use the 'models' or 'find' commands to search for specific model capabilities like vision, tools, or long-context support.

How do I set up the OpenRouter skill in Claude Code?

You need to export your OpenRouter API key as the environment variable SKILL_OPENROUTER_API_KEY and use the uv run command provided in the documentation.

OpenRouter Multi-Model Gateway

Name: OpenRouter Multi-Model Gateway
Author: tsilva

bytsilva

0•

数据科学与机器学习

Integrates 300+ AI models into Claude Code for specialized tasks, high-fidelity image generation, and cross-model reasoning.

The OpenRouter skill extends Claude Code's native capabilities by providing a unified interface to over 300 external AI models from providers like OpenAI, Google, Meta, and Mistral. It allows developers to delegate specific tasks to specialized models—such as using Gemini for long-context analysis or GPT-5 for complex reasoning—while also enabling high-quality image generation with advanced controls for transparency and aspect ratios. By centralizing model access through a single API key, it simplifies multi-model workflows and automated model discovery within the terminal.

主要功能

01Advanced image generation with support for transparency, custom aspect ratios, and 4K resolution.

02Support for sequential model chaining and parallel model comparison workflows.

03Access to 300+ AI models including GPT-5, Gemini 3, Llama, and Mistral through a single API.

04Model discovery tools to find specialized capabilities like vision, long-context, or tool use.

050 GitHub stars

06Zero-setup execution using UV for automatic dependency management and portability.

使用场景

01Benchmarking prompts across multiple LLMs simultaneously to ensure output consistency and quality.

02Generating production-ready UI assets and logos with transparent backgrounds directly in the dev environment.

03Delegating specialized tasks to the most cost-effective or highest-performing model for a specific domain.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tsilva/claude-skills openrouter

For use in Claude.ai and ChatGPT

Download Skill