Skip to content
Get packs

API Gateway Remote MCP Server

Turn Google Cloud API Gateway REST APIs into remote MCP tool servers for AI agents

Source Download skill
Repository
google/skills
Stars
21,782
Licence
Apache-2.0
Files
7 files

What it does

This skill teaches an AI agent how to set up Google Cloud API Gateway so an existing REST API can be used as a set of tools by AI agents, using the Model Context Protocol (MCP) over HTTP. It guides the agent through annotating OpenAPI 3.x specs, securing the tool list with JWTs or API keys, migrating OpenAPI 2.0 specs, connecting Google ADK or API hub, and troubleshooting MCP errors. The feature is in Public Preview.

Great for

  • Backend or platform engineers who already run REST APIs on Google Cloud API Gateway and want agents to call them
  • Teams migrating OpenAPI 2.0 specs to OpenAPI 3.x for MCP tool exposure
  • Developers building agents with Google ADK who need a remote MCP endpoint

Skip it if

  • Apigee X proxy bundles, outbound LLM model routing, or GKE Gateway API, which the skill explicitly says it does not cover
  • Production systems that cannot accept Public Preview limits and changing behaviour
  • Non-technical users without a Google Cloud project and billing account, since deployment requires both

Try asking

  • “Convert my OpenAPI 2.0 spec to OpenAPI 3.1 and annotate the operations as MCP tools on API Gateway”
  • “Secure my MCP tools/list endpoint with a JWT and hide my admin routes from agents”
  • “Help me connect my API Gateway MCP server to a Google ADK McpToolset”
  • “Why am I getting MCP error -32602 from my API Gateway server?”
Step-by-step guide: how to use API Gateway Remote MCP ServerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 7 original files, the Apache-2.0 licence and an attribution note.

    api-gateway-mcp-server.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip api-gateway-mcp-server.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip api-gateway-mcp-server.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Review the validator script's output before deploying, and run it on every OpenAPI spec change
  • Use the most restrictive IAM role for each step and avoid broad project-wide grants
  • Keep all MCP operations in a single OpenAPI document per API config, and check tools/list after deployment to confirm the tools you expect are exposed

What's inside

  • references/adk-and-api-hub-discovery.md
  • references/attribution.md
  • references/mcp-limitations-and-troubleshooting.md
  • references/openapi-agent-tool-optimizer.md
  • references/openapi-v2-to-v3-mcp-migration.md
  • scripts/validate_mcp_openapi.py
  • SKILL.md

Redistributed unmodified from google/skills under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.