Skip to main content

Developer Tools

Last updated on October 6, 2026

Overview​

This module covers tools that help you build with AccelByte Gaming Services (AGS) from wherever you work: your terminal, your scripts, and your AI assistant.

Which tool to use​

The AGS CLI and the AGS API Model Context Protocol (MCP) Server reach the same AGS APIs, and both work with an AI assistant. With either one, your assistant can find the right endpoint, explain what an API does, and look up or change something when you ask in plain language.

The difference is where each tool runs and how predictable its results are:

IfUse
You want to start right away, with nothing to installAGS API MCP Server
Your assistant runs in a web browser, such as Claude or ChatGPT on the web, and can't run commandsAGS API MCP Server
A multi-step change must run the same way every timeAGS CLI workflows
You want to review every call before anything runsAGS CLI workflows
You run in a CI pipeline or script, without anyone logging inAGS CLI

Both tools offer workflows, but they run differently. In the MCP server, your AI assistant follows a workflow's guidance and decides each step, so two runs can differ. An AGS CLI workflow is a file with fixed steps, so the same inputs always produce the same calls, and a dry run shows each one first.

More tools on the way​

These tools are in development or early availability:

  • AI Marketplace Plugin: Connects AI coding assistants such as Claude Code and Codex to the AGS CLI and to the AGS MCP servers. It makes AGS workflows available to an AI assistant as callable tools.
  • AGS Extend SDK MCP Server: An MCP server that gives an AI assistant knowledge of the AGS Extend SDK for Go, C#, Java, and Python.