skills.rest
google
googleOFFICIAL
Explore all skills by this author →
Website𝕏@GoogleOSSGitHub

google-cloud-solution-agentic-ai-data-science-workflow

Designs multi-product agentic data science architectures on Google Cloud with deployment and validation plans.

19.1k|1.5k|Updated Mar 31, 2026
Search Tags:#architecture design#Mermaid#terraform#Google Cloud#Multi-Agent#Agentic AI#Data science
One-click install
npx skills add https://github.com/google/skills --skill google-cloud-solution-agentic-ai-data-science-workflow
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: google-cloud-solution-agentic-ai-data-science-workflow
Source: https://github.com/google/skills/tree/main/skills/cloud/google-cloud-solution-agentic-ai-data-science-workflow
Command: npx skills add https://github.com/google/skills --skill google-cloud-solution-agentic-ai-data-science-workflow

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

Architecting an agent-based data analytics or ML solution on Google Cloud requires choosing the right agent design pattern, mapping components to products, and producing deployment-ready infrastructure, which is complex and error-prone without structured guidance.

Core Features & Use Cases

  • Four-phase guided workflow: Requirements discovery, solution design, implementation planning, and solution validation with explicit user confirmation gates.
  • Product mapping and design recommendations: Maps workload components to Google Cloud products (Cloud Run, Gemini Enterprise Agent Platform, AlloyDB, BigQuery) with grounded best-practice guidance across security, reliability, cost, and performance pillars.
  • Architecture and IaC generation: Produces Mermaid architecture diagrams, a solution-architecture-guide.md document, Terraform code, and validation scripts.
  • Use Case: A team wants to build a multi-agent data science system where a coordinator agent routes analytical queries to database and ML agents; this Skill walks them from requirements gathering through a validated Google Cloud deployment.

Quick Start

Ask the agent to design a Google Cloud architecture for an agentic data science workflow that analyzes data in BigQuery and AlloyDB, and answer its clarifying requirements questions.

Frequently Asked Questions about google-cloud-solution-agentic-ai-data-science-workflow

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I design an agentic data science architecture on Google Cloud?▼

Follow the four-phase workflow: discover requirements through clarifying questions, select an agent design pattern and map components to Google Cloud products, generate Terraform and deployment instructions, then validate the deployment with verification scripts.

Which agent design pattern should I use for data analytics workloads?▼

The coordinator pattern is recommended for most workloads. Use single-agent for simple single-source tasks, sequential or parallel patterns for deterministic pipelines, and review-and-critique for high-stakes tasks needing critic loops.

Cloud Run vs GKE for hosting agent runtimes?▼

Cloud Run is the recommended primary for agent runtimes due to serverless simplicity. GKE offers maximum infrastructure control and stateful pods but adds high operational complexity and cluster management overhead.

When should I not use this data science workflow skill?▼

Avoid it for simple queries, non-agentic pipelines, general cloud reviews, or writing agent code directly. It is scoped to architecting multi-product agent-based analytics or ML solutions, not implementation of agent logic.

Does the workflow require user approval between phases?▼

Yes. The workflow halts for explicit user confirmation after the technical decomposition, solution architecture, and implementation plan before proceeding to the next phase, iterating until approval is given.

Related Skills

Explore More Data & Analytics Skills→
withmartian-sandboxwithmartian-sandbox

ops-migration-planning

Create structured migration plans for infrastructure and workload transitions.

Official
Advanced
googlegoogle
17.1k

google-cloud-global-frontend-configuration

Design and deploy Google Cloud global external Application Load Balancers with Terraform HCL or gcloud CLI scripts.

Official
Advanced
AgentientAgentient
2

gcp-hardening

Enforce IAM least privilege, VPC security, and Secret Manager usage on Google Cloud Platform.

Official
Basic
EdgeHuntEdgeHunt

ck:google-adk-python

Build and deploy AI agents with Google ADK Python.

Official
Advanced