cloudbrew 1.0.0

CloudBrew: vendor-neutral cloud orchestration CLI

CloudBrew

CloudBrew is a next-generation, vendor-neutral cloud orchestration CLI. It abstracts Infrastructure-as-Code (Terraform, Pulumi) behind a unified, intelligent interface that handles routing, policy enforcement, autoscaling, and task offloading.

Stop writing boilerplate HCL or YAML for every resource. Just brew it.

---

Key Features

• Intelligent Routing: Automatically decides whether to provision resources via a Hot Pool (L1), Warm Pool (L2), or perform a Cold Build (L3) based on latency and availability.
• Universal Adapters: Seamlessly wraps Terraform and Pulumi. Switch providers (AWS, Azure, GCP, Noop) with a single flag.
• Dynamic Autoscaling: Built-in logic to scale resources based on custom metrics (e.g., 1:5@cpu:70) with cooldown handling.
• Governance & Policy: Enforce constraints (Allowed Regions, Cost Budgets, Instance Types) via policies.json before infrastructure is touched.
• Stack Blueprints: Deploy complex architectures (e.g., LAMP, K8s clusters) using Jinja2-templated blueprints.
• Drift Detection: Detect manual changes in your cloud environment against your local state.
• async Task Offloading: Heavy provisioning tasks can be offloaded to a background SQLite-backed worker queue.

---

Installation

Prerequisites

• Python 3.9+
• (Optional) Terraform or Pulumi installed on your system PATH.

Install via Pip

pip install cloudbrew

Intelligent Provisioning (Smart Router)

Let the Intelligent Router decide the best way to provision (Hot Cache vs Cold Build) and find the cheapest spot prices.

cloudbrew create-vm worker-node --smart --yes

Dynamic Resource Resolution

CloudBrew uses fuzzy matching to find resources. You don't need to remember exact provider resource names.

# CloudBrew resolves "bucket" to "aws_s3_bucket" or "google_storage_bucket"
cloudbrew bucket my-data-lake --region us-east-1

Deploying Stacks

Deploy multi-resource environments using blueprints.

# List available stacks
cloudbrew stacks

# Deploy a LAMP stack
cloudbrew stack deploy lamp my-web-app --env prod --region us-east-1

Autoscaling

Apply autoscaling policies to your resources using the compact syntax: min:max@metric:threshold.

# Scale 'web-worker' between 1 and 5 instances. 
# Scale up if CPU > 70%, with a 60s cooldown.
cloudbrew create --name web-worker --autoscale "1:5@cpu:70,60" --spec examples/spec.json

Offloading Tasks (Async)

Don't wait for Terraform to finish. Offload the task to a background worker.

# 1. Enqueue the task
cloudbrew create-vm heavy-db --size large --async

# 2. Run the worker (in a separate terminal)
cloudbrew offload run-worker

Drift Detection

Check if your infrastructure has drifted from the known state.

cloudbrew drift web-server-01

Policy Enforcement

CloudBrew checks policies.json before every operation.

{
  "allowed_regions": ["us-east-1", "eu-central-1"],
  "banned_instance_types": ["xlarge", "metal"],
  "max_monthly_cost": 500
}

If you try to create an xlarge instance or deploy to us-west-1, the CLI will block the operation and report a violation

Architecture

CLI Layer (Typer): Parses commands and routes them.

Logic Layer (LCF):

1. Intelligent Router: Optimizes placement.

2. Resource Resolver: Maps intent ("bucket") to implementation.

3. Policy Engine: Validates specs against rules.

Adapter Layer:

1. TerraformAdapter: Generates HCL, manages state, streams output.

2. PulumiAdapter: Uses Automation API.

3. State Layer: SQLite (cloudbrew.db) acts as the state of truth for the orchestration logic.