cloudcosting 1.3.0

Multi-cloud infrastructure cost estimation tool

cloudcosting

Multi-cloud infrastructure cost estimation tool. Fetches real-time pricing from cloud provider APIs (AWS Pricing API), caches results locally, and produces structured YAML/JSON cost breakdowns. Includes a comparison command for side-by-side multi-scenario cost analysis with docsmith-compatible output for Word document generation.

Architecture

Config YAML -> Config Loader -> Estimator -> Provider Registry -> AWS Provider
                                                                    |
                                                              Calculator per resource type
                                                                    |
                                                              Pricing Adapter (API + Cache)
                                                                    |
                                                              Estimate output (YAML/JSON)

Layers

Layer	Module	Responsibility
Domain	domain.py	Dataclasses, exceptions, serialization
Config	config.py	YAML parsing, structural validation
Cache	cache.py	File-based pricing cache with TTL
Estimator	estimator.py	Transaction script: config -> providers -> aggregate
Provider	providers/aws/	AWS-specific pricing adapter and calculators
Formatters	formatters.py	Output format transformations (docsmith)
Composer	composer.py	Multi-scenario comparison document composition
CLI	cli.py	Command-line interface (estimate, cache, compare)

Supported AWS Resource Types

Type	Calculator	Required Params
rds	RDS instances	engine, instance_class, storage_gb
ec2	EC2 instances	instance_type
nat_gateway	NAT Gateways	(none required)
alb	Application Load Balancers	(none required)
ebs	EBS Volumes	size_gb
s3	S3 Storage	size_gb

Installation

pipx install cloudcosting

This makes the cloudcosting command available globally.

Development Setup

For contributing or local development:

cd cloudcosting
uv venv .venv
source .venv/bin/activate
uv pip install -e ".[dev]"

Usage

Estimating Costs

# Run estimation (YAML output)
cloudcosting estimate config.yaml

# Output as JSON
cloudcosting estimate config.yaml --format json

# Output as docsmith-compatible YAML (for Word document generation)
cloudcosting estimate config.yaml --format docsmith -o estimate.yaml

# Pipe directly to docsmith for Word document
cloudcosting estimate config.yaml --format docsmith | docsmith -

# Write to file
cloudcosting estimate config.yaml -o costs.yaml

# Use a specific AWS credentials profile
cloudcosting estimate config.yaml --profile production

Comparing Scenarios

Compare costs across multiple infrastructure configurations. Produces a docsmith-compatible YAML document with side-by-side cost tables.

# Compare two configurations
cloudcosting compare Small:small.yaml Large:large.yaml

# With custom document title and line-item detail
cloudcosting compare Small:small.yaml Large:large.yaml \
  --title "Project Phoenix" --detail

# Write comparison to file, then generate Word document
cloudcosting compare Small:small.yaml Large:large.yaml -o comparison.yaml
docsmith comparison.yaml

# Bare paths (scenario names derived from filenames)
cloudcosting compare small.yaml large.yaml

# Multiple scenarios with a shared AWS profile
cloudcosting compare Small:s.yaml Medium:m.yaml Large:l.yaml --profile production

Scenario specs use the format Name:path or just path. When the name is omitted, the filename stem is used (e.g., small.yaml becomes scenario name small).

Tip: Label your resources in config files for meaningful comparisons. Resources are aligned across scenarios by their label field. Without labels, resources get auto-generated names like EC2 t3.micro which may not match as expected across different configurations.

Cache Management

cloudcosting cache status
cloudcosting cache refresh aws
cloudcosting cache refresh

All commands can also be run via python -m cloudcosting (e.g., python -m cloudcosting estimate config.yaml).

Output Formats

Format	Flag	Description
yaml	--format yaml (default)	Structured estimate with full metadata
json	--format json	Same structure as YAML, serialized as JSON
docsmith	--format docsmith	docsmith-compatible YAML for Word document generation

The compare command always produces docsmith-compatible YAML output.

Example Config

provider: aws
region: us-east-1

resources:
  - type: rds
    label: Primary Database
    engine: postgres
    instance_class: db.r6g.xlarge
    storage_gb: 250
    multi_az: true

  - type: ec2
    label: Web Servers
    instance_type: t3.micro
    count: 3

  - type: nat_gateway
    label: NAT Gateways
    count: 2

  - type: alb
    label: Application Load Balancer

  - type: ebs
    label: Data Volumes
    size_gb: 500
    volume_type: gp3
    count: 3

  - type: s3
    label: Document Storage
    size_gb: 1000

Testing

# Run all tests
pytest tests/ -v

# Run specific test module
pytest tests/unit/test_domain.py -v
pytest tests/unit/test_composer.py -v
pytest tests/unit/providers/aws/test_rds.py -v

79 unit tests covering domain invariants, config validation, cache behavior, calculator arithmetic, full estimation pipeline, and comparison composition.

Adding New Resource Types

1. Create a calculator module in providers/aws/calculators/ with validate() and estimate() functions
2. Register it in providers/aws/provider.py CALCULATOR_REGISTRY
3. Add tests in tests/unit/providers/aws/

Adding New Providers

1. Create a provider package under providers/ (e.g., providers/azure/)
2. Implement the same interface as AwsProvider (with estimate_resources())
3. Register it in providers/registry.py