claude-starter

🥈Silver

Claude-starter is a platform-agnostic framework for production-ready Claude Code configurations. It includes 40+ skills, meta-commands, and workflow automation tools. Operations teams use it to streamline AI integration across projects, reducing setup time and ensuring consistency.

7450Updated 1mo ago

Intermediate30min to implementautomation

Saves ~180 min per use

Quick InstallView Source

git clone https://github.com/raintree-technology/claude-starter.git

Works with:

Claude

Overview

About This Skill

How to Use

[{"step":"Install the claude-starter CLI tool","action":"Run `pip install claude-starter` or use the provided Docker image `ghcr.io/claude-starter/cli:latest`. Verify installation with `claude-starter --version`.","tip":"Ensure you have Python 3.10+ installed. For Docker-based setups, pull the latest image to avoid version mismatches."},{"step":"Initialize a new project configuration","action":"Run `claude-starter init --project-name [PROJECT_NAME] --template python-docker-k8s` to scaffold a production-ready setup. Replace `[PROJECT_NAME]` with your project’s name.","tip":"Use `--template` to select a pre-configured template (e.g., `python-docker-k8s`, `nodejs-serverless`, or `go-microservice`). Templates include common skills and meta-commands for the tech stack."},{"step":"Customize the configuration","action":"Edit the generated `claude-config.json` file to match your project’s requirements. Update dependencies, environment variables, and error-handling rules. For example, modify the `python` skill’s `dependencies` array to include your project’s libraries.","tip":"Use the claude-starter documentation to explore available skills and meta-commands. Add custom skills by referencing their GitHub repositories (e.g., `claude-starter/skills/aws` for AWS integrations)."},{"step":"Test the configuration locally","action":"Run `claude-starter validate` to check for syntax errors or missing dependencies. Then, use meta-commands like `claude dev` to start a local development server or `claude test` to run tests.","tip":"Enable verbose logging with `claude-starter validate --debug` to troubleshoot issues. Use `claude lint` to catch style or type errors before deployment."},{"step":"Deploy to your target environment","action":"Use the `deploy` meta-command for your environment (e.g., `claude deploy staging`). For Kubernetes, ensure your cluster is configured and your `kubectl` context is set to the correct namespace.","tip":"For production deployments, use `claude deploy prod --dry-run` to preview changes before applying. Monitor deployments with `kubectl logs` or your cloud provider’s dashboard."}]

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/raintree-technology/claude-starter

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Generate a production-ready Claude Code configuration for [PROJECT_NAME] using the claude-starter framework. Include: 1) A [SKILL_SET] skill configuration, 2) [NUMBER] meta-commands for common operations, 3) Error handling for [SPECIFIC_ERROR_CASE], and 4) A deployment-ready [ENVIRONMENT_TYPE] setup. Use the claude-starter CLI tool to scaffold this configuration.

Example Output

Here’s a production-ready Claude Code configuration for **Project Phoenix**, a Python-based microservice for real-time fraud detection. The configuration leverages the claude-starter framework to ensure consistency across environments and includes the following components:

### **Core Configuration (`claude-config.json`)**
```json
{
  "project_name": "Project Phoenix",
  "version": "1.0.0",
  "skills": {
    "claude-starter/skills/python": {
      "python_version": "3.11",
      "dependencies": ["fastapi==0.109.0", "uvicorn==0.27.0", "pydantic==2.5.3"],
      "linting": true,
      "testing": {
        "framework": "pytest",
        "coverage_threshold": 90
      }
    },
    "claude-starter/skills/docker": {
      "base_image": "python:3.11-slim",
      "multi_stage_build": true,
      "optimized_layers": true
    },
    "claude-starter/skills/kubernetes": {
      "namespace": "fraud-detection",
      "resource_limits": {
        "cpu": "1000m",
        "memory": "512Mi"
      }
    }
  },
  "meta_commands": {
    "dev": {
      "description": "Start the development server with hot-reload",
      "command": "uvicorn main:app --reload --host 0.0.0.0 --port 8000"
    },
    "test": {
      "description": "Run all tests with coverage",
      "command": "pytest --cov=app --cov-report=term-missing"
    },
    "lint": {
      "description": "Run static analysis and linting",
      "command": "ruff check . && mypy ."
    },
    "deploy": {
      "description": "Deploy to staging environment",
      "command": "kubectl apply -f k8s/staging/ && kubectl rollout status deployment/fraud-detection -n fraud-detection"
    }
  },
  "error_handling": {
    "timeout": 30,
    "retry_attempts": 3,
    "fallback_commands": ["claude-starter/skills/python/debug"],
    "specific_cases": {
      "database_connection_failure": {
        "action": "claude-starter/skills/python/retry_db_connection",
        "message": "Database connection failed. Retrying in 5 seconds..."
      }
    }
  },
  "environments": {
    "dev": {
      "docker_compose": "docker-compose.dev.yml",
      "env_file": ".env.dev"
    },
    "staging": {
      "k8s_manifests": "k8s/staging/",
      "secrets": "k8s/secrets/staging-secrets.yaml"
    },
    "prod": {
      "k8s_manifests": "k8s/prod/",
      "secrets": "k8s/secrets/prod-secrets.yaml",
      "monitoring": {
        "prometheus": true,
        "grafana_dashboard": "fraud-detection-prod"
      }
    }
  }
}
```

### **Key Features**
1. **Skill Integration**: Pre-configured skills for Python, Docker, and Kubernetes ensure consistency across environments. The Python skill includes linting, testing, and dependency management, while the Docker skill optimizes multi-stage builds for production.
2. **Meta-Commands**: Four meta-commands (`dev`, `test`, `lint`, `deploy`) streamline common operations. For example, running `claude dev` starts the FastAPI server with hot-reload, while `claude test` executes the test suite with coverage reporting.
3. **Error Handling**: The configuration includes a 30-second timeout with 3 retry attempts for transient failures. For database connection issues, it automatically triggers a retry with a 5-second delay and logs the attempt.
4. **Environment-Specific Setups**: The `dev` environment uses Docker Compose for local development, while `staging` and `prod` leverage Kubernetes manifests. The `prod` environment includes Prometheus monitoring and Grafana dashboards for observability.

This configuration reduces setup time by 70% compared to manual configurations and ensures all environments are reproducible. The claude-starter framework’s modular design allows for easy customization—e.g., adding a new skill like `claude-starter/skills/gcp` for cloud deployments—without disrupting existing workflows.