Guide · Agent skill
How to use the Borderless Data Lakehouse Designer agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
This guide explains how to install and use the Borderless Data Lakehouse Designer, an agent skill that guides an AI agent through designing a data platform spanning Google Cloud, AWS S3 and Azure Blob Storage with Apache Iceberg tables and BigQuery data agents. It is aimed at architects and data engineers planning a multi-cloud lakehouse. By the end you will know how to install the skill, how to phrase a request that triggers it, and what design outputs to expect.
Borderless Data Lakehouse Designergoogle/skills · Apache-2.0 Download the skill Quick start
- 1Download and unzip. Download the skill ZIP from this page and unzip it so the folder contains SKILL.md and its references and assets subfolders.
- 2Place the skill folder. Move the unzipped folder into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ folder for one project.
- 3Or upload in the Claude apps. In the Claude apps, open Settings, go to Capabilities, and upload the skill under Skills.
- 4Describe your multi-cloud platform. Ask in plain language for a borderless lakehouse design across your clouds, naming the data sources and goals.
- 5Answer discovery questions. Respond to the agent's questions about data sources, metadata, security, analytics and expected queries.
- 6Confirm the decomposition. Review the technical decomposition and ask for changes until it matches your workload.
- 7Review the architecture guide. Check the product mapping, Mermaid diagram and solution-architecture-guide.md, then request revisions or approval.
What an agent skill is and what this one does
An agent skill is a folder of instructions (centred on a SKILL.md file) that an AI agent loads when a request matches the skill's description. This one, named google-cloud-solution-agentic-ai-borderless-data-lakehouse, is built for designing a multi-cloud data lakehouse rather than running one.
Once installed, the agent asks discovery questions about your data sources, security and credential requirements, analytical workloads and the natural language queries AI agents or end users will run. It then identifies components, proposes Google Cloud products, draws a Mermaid architecture diagram and compiles a solution architecture guide, asking for your confirmation at each stage. The skill also describes implementation and validation phases, but the design phase is the main focus.
- Requirements discovery and component decomposition, confirmed by you before moving on
- Product mapping and a Mermaid diagram separating the ingestion and serving subsystems
- Design recommendations on IAM, credential vending and least privilege
- A solution-architecture-guide.md file based on the included output template
Install the skill from the ZIP download
Download the ZIP for this skill from this page and unzip it. The folder should contain SKILL.md, the references/ and assets/ subfolders. Then place it where your agent looks for skills: your user-level folder ~/.claude/skills/ to make it available everywhere, or a project's .claude/skills/ folder to limit it to one project.
If you use the Claude apps instead, open Settings, go to Capabilities, and upload the skill under Skills.
- Personal install: unzip into ~/.claude/skills/
- Project install: unzip into the project's .claude/skills/ folder
- Claude apps: Settings → Capabilities → Skills, then upload
Trigger the skill with a natural request
You do not need a special command. Describe a multi-cloud data platform in plain language and the agent should match the skill's description. The description mentions Cloud Storage, AWS S3, Azure Blob, Cross-Cloud Interconnect, BigQuery data agents and Gemini Enterprise Agent Platform, so requests using those terms are the most likely to trigger it.
Example requests that fit the skill:
- Design a borderless lakehouse using Iceberg tables where raw data stays in S3 and Azure Blob, and BigQuery serves analysts.
- Help me architect multi-cloud storage for Cloud Storage, AWS S3 and Azure Blob with a shared catalog.
- I need a design doc for an agentic AI data platform with BigQuery data agents, reviewed before any infrastructure is built.
Work through the design conversation
Expect the agent to ask questions before it proposes anything. Answer what you know and say when you do not know. Each stage ends with a request for confirmation: first the technical decomposition, then the solution architecture. Ask for changes at either point and the agent regenerates that output.
The design phase fetches Google Cloud architecture documentation during its work, so the result is only as current as those pages and your own review of them.
- Discovery: data sources, metadata management, security and credentials, analytical needs, expected queries
- Decomposition: components, including any on-prem or other-cloud parts you must integrate with
- Architecture: product mapping, Mermaid diagram, design recommendations, then the guide file
Safety notes: network use and commands
The skill uses the network, because it retrieves Google Cloud documentation pages. It is listed as not running scripts. Its design phase does not require gcloud, REST API or MCP tools. Those are needed only if you run the implementation phase.
If you do run implementation, the skill's attribution instructions require every gcloud command to be prefixed with a CLOUDSDK_METRICS_REQUEST_ATTRIBUTION environment variable set inline, and warn against using gcloud config set for this. Review any command the agent proposes before running it against a real project.
- No scripts are run by the skill
- Network access is used to fetch the documentation it cites
- Implementation tooling is needed only for the implementation phase
CLOUDSDK_METRICS_REQUEST_ATTRIBUTION="agent-skills/1.0 (skill:google-cloud-solution-agentic-ai-borderless-data-lakehouse)" \
gcloud <command> [flags]Tips for getting a useful design
Bring your real constraints: which clouds already hold data, where credentials live, and which regions or compliance rules apply. The discovery questions are built around these, and vague answers produce vague designs.
Treat the output as a draft for review. Check the product choices against current Google Cloud documentation before you present the guide to others.
- List every data source and its cloud up front
- State whether a single-cloud warehouse would already meet the need, since the skill is not meant for that case
- Use the guide as a review checklist with your team rather than a final sign-off
What people use it for
- Planning where Iceberg tables live across Cloud Storage, AWS S3 and Azure Blob and how BigQuery reads them
- Preparing a design document for an architecture review before any infrastructure is built
- Mapping ingestion and serving subsystems, with Managed Service for Apache Spark as shared ETL component, for a team to critique
- Checking security choices such as IAM, credential vending and least privilege against a multi-cloud plan
Frequently asked questions
How do I install the Borderless Data Lakehouse Designer skill?
Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use or a project's .claude/skills/ folder for one project. In the Claude apps, you can instead upload it under Settings, then Capabilities, then Skills.
Is this skill free to use?
The skill is published under the Apache-2.0 licence, as listed on this page. Check the licence file in the source repository for the full terms.
What does the skill need to run?
The design phase needs search or fetch tools to read Google Cloud documentation and Mermaid for the diagram. gcloud, REST API or MCP tooling is needed only if you run the implementation phase.
Does the skill run scripts or access the network?
It does not run scripts, according to this page. It does use the network to retrieve the Google Cloud documentation pages it cites.
Can I use this skill for a single-cloud data warehouse?
No. The skill says not to use it for single-cloud data warehouses. For Knowledge Catalog governance or Spark IDE analytics, the skill points to google-cloud-solution-agentic-analytics-spark-knowledge-catalog instead.
Why does the agent keep asking questions before designing?
The workflow requires requirements discovery and asks you to confirm the technical decomposition before designing. Answering the questions with specific details gives a more accurate architecture.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
This skill suits architects and data engineers who want a structured, confirm-as-you-go checklist for a multi-cloud Iceberg and BigQuery lakehouse design before building anything. It is not suitable for single-cloud warehouses or for work centred on Knowledge Catalog governance or Spark IDE analytics. Treat its output as a draft to verify against current Google Cloud documentation.
Download the skill