How does it handle security and rate limiting?

It utilizes AgentgatewayPolicy resources to manage authentication via Kubernetes secrets, CORS configuration, and sophisticated rate limiting for both requests and tokens.

Which LLM providers are supported by this skill?

The skill supports a wide range of providers including Ollama (local), OpenAI, Anthropic, Google Gemini/Vertex AI, Azure OpenAI, and AWS Bedrock.

What is agentgateway?

agentgateway is a Linux Foundation project that extends the Kubernetes Gateway API specifically for AI and agentic workloads, providing unified routing for LLMs and MCP servers.

Does it help with upgrading the gateway?

Absolutely. The skill includes commands like /gw-versions and /gw-upgrade to help you compare installed versions against the latest releases and guide you through safe Helm upgrades.

Can I use this for local development?

Yes, it specifically supports routing to local Ollama instances (often via host.internal), making it ideal for hybrid local and cloud AI development.

AgentGateway Manager

Name: AgentGateway Manager
Author: totallyGreg

bytotallyGreg

•

クラウドインフラストラクチャ

Configures and manages AI/LLM routing, MCP server connections, and traffic policies on Kubernetes using agentgateway.

The AgentGateway Manager skill provides expert guidance for implementing and maintaining agentgateway, a specialized extension of the Kubernetes Gateway API designed for AI workloads. It enables developers to streamline LLM provider routing across OpenAI, Anthropic, and local Ollama instances, while also managing Model Context Protocol (MCP) endpoints and external processing. With built-in diagnostic commands and YAML generation, this skill simplifies complex tasks like setting up token-based rate limiting, configuring cloud backends with secrets, and performing validated Helm upgrades to keep your AI infrastructure secure and up to date.

主な機能

01Automated Kubernetes YAML generation for backends and security policies

021 GitHub stars

03Advanced traffic policies including token-based rate limiting and CORS

04Model Context Protocol (MCP) server routing via HTTP/SSE endpoints

05Multi-provider LLM routing for OpenAI, Anthropic, Gemini, Vertex AI, Azure, and Bedrock

06Integrated diagnostic tools for status monitoring, log tailing, and version validation

ユースケース

01Implementing Model Context Protocol (MCP) routing for agentic workflows on Kubernetes

02Setting up a hybrid AI environment routing between local Ollama and cloud-based LLM providers

03Managing the lifecycle of agentgateway through automated version checks and Helm upgrade guidance

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add totallygreg/claude-mp agentgateway

For use in Claude.ai and ChatGPT