[New Skill]: Google Cloud Well-Architected Framework: Sustainability Pillar

[feudadrien]

Feature Request: New Agent Skill for Google Cloud Well-Architected Framework: Sustainability Pillar

Description

This request is for the addition of a new Agent Skill covering the Sustainability pillar of the Google Cloud Well-Architected Framework. Currently, the repository includes skills for Security, Reliability, and Cost Optimization, but Sustainability is a key pillar that is missing.

Proposed Skill Details

• Name: google-cloud-waf-sustainability
• Source Material: Google Cloud Well-Architected Framework: Sustainability
• Structure: Consistent with existing WAF skills (Security, Reliability, Cost Optimization).

Draft Content

I have prepared a draft of the SKILL.md file based on the official documentation and the style of existing skills in this repository.

SKILL.md Preview:

---
name: google-cloud-waf-sustainability
description: >-
    Generates sustainability-focused guidance for Google Cloud workloads
    based on the design principles and recommendations in the Google Cloud
    Well-Architected Framework. Use this skill to evaluate a workload, identify
    sustainability requirements, and provide actionable recommendations for
    building and managing energy-efficient and low-carbon workloads in Google Cloud.
---
# Google Cloud Well-Architected Framework skill for the Sustainability pillar
## Overview
The Sustainability pillar of the Google Cloud Well-Architected Framework provides principles and recommendations to help you minimize the environmental impact of your cloud workloads. A sustainable architecture is resilient and optimized, creating a positive feedback loop of higher efficiency, lower cost, and lower environmental impact. Sustainability in the cloud is a shared responsibility between Google (sustainability *of* the cloud) and you, the customer (sustainability *in* the cloud).
## Core principles
The recommendations in the sustainability pillar of the Well-Architected Framework are aligned with the following core principles:
- **Use regions that consume low-carbon energy**: Select Google Cloud regions with the highest Carbon-Free Energy (CFE%) and the lowest grid carbon intensity. Use tools like the Region Picker and enforce location policies.
  - [Grounding document](https://docs.cloud.google.com/architecture/framework/sustainability/low-carbon-regions)
- **Optimize AI and ML workloads for energy efficiency**: Choose efficient models, specialized hardware (GPUs/TPUs), and optimize training and inference processes to reduce the carbon footprint of AI.
  - [Grounding document](https://docs.cloud.google.com/architecture/framework/sustainability/ai-ml-energy-efficiency)
- **Optimize resource usage for sustainability**: Right-size resources, use serverless services that scale to zero, and leverage the latest energy-efficient processor architectures.
  - [Grounding document](https://docs.cloud.google.com/architecture/framework/sustainability/optimize-resource-usage)
- **Develop energy-efficient software**: Write code that minimizes CPU, memory, and network usage. Use efficient algorithms, data structures, and event-driven patterns.
  - [Grounding document](https://docs.cloud.google.com/architecture/framework/sustainability/energy-efficient-software)
- **Optimize data and storage for sustainability**: Minimize data sprawl, use appropriate storage classes, and automate lifecycle management to reduce the energy cost of data storage.
  - [Grounding document](https://docs.cloud.google.com/architecture/framework/sustainability/optimize-storage)
- **Continuously measure and improve sustainability**: Use the Carbon Footprint tool to quantify emissions, identify hotspots, and track the impact of optimization efforts over time.
  - [Grounding document](https://docs.cloud.google.com/architecture/framework/sustainability/continuously-measure-improve)
## Relevant Google Cloud products
The following are examples of Google Cloud products and features that are relevant to sustainability:
- **Visibility and monitoring**:
  - **Carbon Footprint**: Native dashboard for visualizing cloud-related emissions.
  - **BigQuery Carbon Export**: Enables granular analysis and custom reporting of carbon data.
  - **Cloud Monitoring and Logging**: Essential for observability and performance optimization.
- **Automation and optimization tools**:
  - **Active Assist / Recommender**: Identifies right-sizing opportunities and unattended projects.
  - **Organization Policy Service**: Enforces resource location constraints to prioritize low-carbon regions.
- **Efficient infrastructure**:
  - **Compute Engine**: Use of energy-efficient machine types (e.g., C4, C4A) and Spot VMs.
  - **Cloud Run**: Serverless environment that scales instances to zero when idle.
  - **GKE (Google Kubernetes Engine)**: Autopilot mode, bin packing, and HPA/VPA for optimal resource utilization.
- **Storage and data**:
  - **Cloud Storage Lifecycle Policies**: Automatically moves data to lower-carbon storage classes (e.g., Archive).
  - **BigQuery**: Efficient data processing and colocation with compute.
## Workload assessment questions
Ask appropriate questions to understand the sustainability-related requirements and constraints of the workload. Choose questions from the following list:
- **Location selection**:
  - How do you incorporate carbon intensity and CFE% metrics into your region selection strategy?
  - Do you use Organization Policies to restrict resource creation to low-carbon regions?
- **Resource optimization**:
  - Are you using serverless or managed services (like Cloud Run or GKE Autopilot) to minimize idle resources?
  - How frequently do you audit and right-size your compute instances based on actual utilization?
  - Do you have a process for identifying and decommissioning unattended projects and resources?
- **Software and Data Efficiency**:
  - How do you optimize your application code to minimize CPU and network demand?
  - What is your strategy for managing data lifecycles and archiving obsolete datasets?
  - Are you using energy-efficient storage formats (like Parquet or Avro) for analytics workloads?
- **Measurement and Governance**:
  - Do you have regular visibility into your Carbon Footprint data?
  - Have you established carbon-related KPIs (e.g., carbon intensity per transaction)?
  - Is sustainability training provided to your architecture and development teams?
## Validation checklist
Use the following checklist to evaluate the architecture's alignment with sustainability recommendations:
- **Region Strategy**: Workloads are deployed in regions with a high CFE% or low CO2 indicator where compliance and latency allow.
- **Scale-to-Zero**: Serverless or auto-scaling configurations are used to eliminate energy consumption by idle resources.
- **Right-sizing**: CPU and memory allocations are regularly adjusted based on Active Assist recommendations.
- **Processor Efficiency**: The latest energy-efficient machine types (e.g., Arm-based C4A) are preferred for new deployments.
- **Storage Lifecycle**: Automated lifecycle rules are active for all major storage buckets and datasets.
- **Data Minimization**: Duplicate, shadow, and dark data are identified and removed regularly.
- **Carbon Visibility**: Carbon Footprint data is exported to BigQuery and reviewed as part of regular operational reviews.
- **Carbon-Aware Scheduling**: Non-urgent batch workloads are scheduled during periods of higher carbon-free energy availability.

Benefits

Adding this skill will enable AI agents to provide comprehensive architectural guidance that aligns with Google's commitment to sustainability and helps customers build more efficient, low-carbon cloud solutions.

[kumar-dhanagopal]

Hi Adrien,

Thank you for the interest & for drafting the skill! I'll loop you into an internal doc that tracks this.