Cloud Run

Cloud Run Skill

Overview

This skill provides access to the Google Cloud Platform Cloud Run MCP server with progressive disclosure for optimal context usage.

Context Savings: ~90% reduction

• MCP Mode: ~15,000 tokens always loaded (7 tools + prompts)
• Skill Mode: ~300 tokens metadata + on-demand loading

Requirements

• Node.js 18+ OR Docker installed and running
• Google Cloud credentials configured via gcloud auth application-default login
• Optional: GOOGLE_CLOUD_PROJECT and GOOGLE_CLOUD_REGION environment variables

Tools

The server provides 8 tools across deployment and management categories:

Deployment Tools

Service Management Tools

Project Management Tools (Local Only)

Quick Reference

# List available tools
python executor.py --list

# List Cloud Run services
python executor.py --tool list-services --args '{"projectId": "my-project", "region": "us-central1"}'

# Get service details
python executor.py --tool get-service --args '{"projectId": "my-project", "region": "us-central1", "serviceName": "my-service"}'

# Get service logs
python executor.py --tool get-service-log --args '{"projectId": "my-project", "region": "us-central1", "serviceName": "my-service"}'

# Deploy file contents
python executor.py --tool deploy-file-contents --args '{"projectId": "my-project", "region": "us-central1", "serviceName": "my-service", "files": [{"path": "main.py", "contents": "..."}]}'

# List GCP projects
python executor.py --tool list-projects

Configuration

Environment Variables

Setup

1. Authenticate with GCP:

   gcloud auth application-default login
   

2. Set default project (optional):

   export GOOGLE_CLOUD_PROJECT=my-project-id
   export GOOGLE_CLOUD_REGION=us-central1
   

3. Use the skill:

   python .claude/skills/cloud-run/executor.py --list
   

Prompts

The MCP server also provides natural language prompts:

Tool Details

deploy-file-contents

Deploy files to Cloud Run by providing their contents directly.

Parameters:

• projectId (string, required): GCP project ID
• region (string, required): Cloud Run region
• serviceName (string, required): Name for the Cloud Run service
• files (array, required): Array of file objects with path and contents

Example:

{
  "projectId": "my-project",
  "region": "us-central1",
  "serviceName": "my-api",
  "files": [
    { "path": "main.py", "contents": "from flask import Flask\napp = Flask(__name__)\n..." },
    { "path": "requirements.txt", "contents": "flask==2.0.0\ngunicorn==20.1.0" }
  ]
}

list-services

List Cloud Run services in a given project and region.

Parameters:

• projectId (string, required): GCP project ID
• region (string, optional): Cloud Run region (defaults to configured region)

get-service

Get detailed information about a specific Cloud Run service.

Parameters:

• projectId (string, required): GCP project ID
• region (string, required): Cloud Run region
• serviceName (string, required): Cloud Run service name

get-service-log

Get logs and error messages for a specific Cloud Run service.

Parameters:

• projectId (string, required): GCP project ID
• region (string, required): Cloud Run region
• serviceName (string, required): Cloud Run service name
• limit (number, optional): Maximum number of log entries

deploy-local-folder

Deploy a local folder to a Cloud Run service (local mode only).

Parameters:

• projectId (string, required): GCP project ID
• region (string, required): Cloud Run region
• serviceName (string, required): Cloud Run service name
• folderPath (string, required): Path to local folder to deploy

list-projects

List available GCP projects (local mode only).

Parameters: None

create-project

Create a new GCP project and attach it to billing (local mode only).

Parameters:

• projectId (string, required): Desired project ID
• projectName (string, optional): Human-readable project name

Integration with Agents

This skill integrates with the following agents:

• devops: For Cloud Run deployments and infrastructure management
• developer: For deploying applications during development
• architect: For service architecture decisions

Troubleshooting