skills.rest
Eigenwise
EigenwiseCOMMUNITY
Explore all skills by this author →
Website𝕏@Kenny_VGitHub

model-gateway

Configure and troubleshoot a local gateway exposing ChatGPT/Codex models in Claude Code.

15|3|Updated Jun 13, 2026
Search Tags:#authentication#routing#claude code#codex#diagnostics#gateway#model picker
One-click install
npx skills add https://github.com/Eigenwise/eigenwise-toolshed --skill model-gateway
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: model-gateway
Source: https://github.com/Eigenwise/eigenwise-toolshed/tree/main/plugins/model-gateway/skills/model-gateway
Command: npx skills add https://github.com/Eigenwise/eigenwise-toolshed --skill model-gateway

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you set up and maintain a local gateway that lets Claude Code access ChatGPT/Codex subscription models through the model picker, while preserving normal Claude model access.

Core Features & Use Cases

  • Gateway Setup and Login: Install, authenticate, and start the local proxy and shim needed for model routing.
  • Model Discovery and Routing: Expose Codex-backed models in Claude Code and route requests to the correct backend family.
  • Day-2 Operations and Recovery: Diagnose ports, auth, discovery, wiring mode, and restore service after failures or upgrades.
  • Workspace and Scope Management: Choose global or per-project wiring, update saved settings, and manage model pins and compatibility mode.

Quick Start

Ask Claude to set up or diagnose the local model gateway for Claude Code and guide the user through the correct wiring mode, login, and restart steps if needed.

Frequently Asked Questions about model-gateway

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

FAQPage Schema
How do I use ChatGPT models inside Claude Code?▼

You can use ChatGPT models inside Claude Code by setting up a local gateway that exposes ChatGPT and Codex models in the model picker, while preserving normal Claude model access.

How do I set up a local proxy to route Claude Code to ChatGPT?▼

To route Claude Code to ChatGPT, you install, authenticate, and start a local proxy and shim, then select the correct wiring mode for your global or per-project scope.

Can I configure Claude Code to access Codex subscription models?▼

Yes, you can access Codex subscription models in Claude Code by configuring a local gateway that handles model discovery and routes requests to the correct backend family.

Why does Claude Code model picker not show ChatGPT models after restart?▼

If the model picker does not show ChatGPT models after a restart, you need to run day-2 diagnostics to check ports, authentication, discovery, and settings wiring to restore service.

How do I manage project-scope versus user-scope settings for model routing?▼

You manage project-scope versus user-scope settings by choosing global or per-project wiring, updating saved settings, and managing model pins and compatibility mode within your environment.

What is the best way to troubleshoot local gateway connection issues in Claude Code?▼

The best way to troubleshoot local gateway connection issues in Claude Code is to run diagnostics for ports, authentication, discovery, wiring mode, and restart timing to restore service after failures.

Related Skills

Explore More Software Engineering Skills→
EigenwiseEigenwise
15

verify

Verify model-gateway CLI and HTTP shim behavior in isolated environments.

Community
Advanced
winsorllcwinsorllc

port-scanner

Scan TCP ports on specified hosts for open services.

Community
Basic
CowboyLogicCowboyLogic

client-config-claudecode

Configure Claude Code user settings with validation and confirmation.

Official
Advanced
obulaiobulai
1

obul-proxy

Proxy HTTP requests through the Obul gateway with x402 payment negotiation.

Official
Intermediate