Snowflake Semantic Views skill: build and validate semantic views with Claude

Makes your AI assistant create, alter and validate Snowflake semantic views with the Snowflake CLI.

Installs: 9K; on skills.sh
Stars: 40.7K; on GitHub
Updated: Oct 2026; last repo commit

snowflake-semanticview has your assistant draft semantic view DDL, check it against your Snowflake account with the snow CLI, then apply it and run a sample query. It confirms the database, schema, role, warehouse and view name first, and validates under a temporary name such as __tmp_validate. It needs Snowflake CLI installed and a connection set up with snow connection add. It is a community skill in awesome-copilot, not one published by Snowflake.

Good for

Snowflake developers building a semantic layer with CREATE SEMANTIC VIEW
Analytics engineers who want DDL checked against Snowflake before it is final
Anyone who needs help installing Snowflake CLI and adding a connection

Install snowflake-semanticview

Claude Code

1
Download the skill folder from https://github.com/github/awesome-copilot/tree/HEAD/skills/snowflake-semanticview.
2
Put it at ~/.claude/skills/snowflake-semanticview/ so that SKILL.md sits directly inside.
3
Start a new session; the skill loads when its description matches the task.

Requires Snowflake CLI (snow).

Install Snowflake CLI first and check it with snow --help; the SKILL.md links to Snowflake's install guide. Add a connection with snow connection add.

Built from the package data we hold. Not yet tested by us.

What's in the skill

When it loads

Create, alter, and validate Snowflake semantic views using Snowflake CLI (snow). Use when asked to build or troubleshoot semantic views/semantic layer definitions with CREATE/ALTER SEMANTIC VIEW, to validate semantic-view DDL against Snowflake via CLI, or to guide Snowflake CLI installation and connection setup.

Sections in SKILL.md

One-Time Setup
Workflow For Each Semantic View Request
Synonyms And Comments (Required)
Validation Pattern (Required)
Example CLI Validation (Template)
Notes

Key facts

Author: (awesome-copilot) community repo; no author named in the skill folder
Repo: github/awesome-copilot
License: MIT
Installs on skills.sh: 9,016 as of 2026-11-02
GitHub stars: 40,666 for the whole github/awesome-copilot repo, 2026-11-02
Last repo commit: 2026-10-30
What is inside: 1 file, SKILL.md; it bundles no scripts
Needs: Snowflake CLI (snow) and a connection added with snow connection add

What it includes

Check snow --help, then add a connection
Verifies the CLI with `snow --help`, points to the Snowflake install docs if it is missing, and sets up a connection with `snow connection add`.
Runs code · One-time setup
Database, schema, role, warehouse, view name
Step 1 of 11. Also confirms that the model follows a star schema.
Try: “Create a Snowflake semantic view over my orders and customers tables in ANALYTICS.PUBLIC.”
Read · Confirm the target
DISTINCT and LIMIT queries, at most 1000 rows
Runs SELECT statements to find how fact and dimension tables relate and to write better comments and synonyms.
Runs code · Discover with SELECT
Runs the DDL with snow sql first
Appends a name like __tmp_validate, executes the DDL with `snow sql` on your configured connection, and loops until it succeeds.
Runs code · Validate under a temporary name
Creates or alters the real semantic view
Runs the validated DDL under the real name, then a sample query with SEMANTIC_VIEW(...).
Runs code · Apply the final DDL
Drops the temporary view
Step 11 removes the temporary semantic view created for validation.
Delete · Clean up
Required, never invented without approval
Uses existing Snowflake comments first. If they are missing, it asks whether it may create them, whether you will write them, or whether it should draft suggestions for you to approve.
Read · Synonyms and comments

From the official docs

Permissions

Medium risk

The skill runs SQL on your Snowflake account, including creating and altering a semantic view. Use a development role and database first.

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

It creates and alters objects in your account
Steps 6 to 11 create a temporary semantic view, apply a create or alter on the real name, and clean up the temporary one. Run it with a role that can only touch the schema you intend. github.com
SKILL.md has no approval step before the final apply
Step 1 confirms the target names, but step 9 applies the final DDL without a separate confirmation. Tell your assistant to show the DDL and wait for you first. github.com
Query results reach your assistant
Step 5 runs SELECT statements to inspect the data. Your assistant reads the rows it gets back to write comments and synonyms. github.com

Technical details

Turning on read-only	No read-only mode. The workflow ends with a create or alter.
Talks to	Snowflake, through the Snowflake CLI and your configured connection
Sign-in and keys	Uses whatever connection you set up with snow connection add.
Scopes	Runs snow sql on the connection you configure Creates a temporary semantic view, then creates or alters the real one Runs SELECT queries with up to 1000 rows to learn about your tables

Questions

What does snowflake-semanticview do?

It walks your assistant through building a Snowflake semantic view: confirm the target, draft the DDL, add synonyms and comments, validate through Snowflake CLI, apply the final DDL, run a sample query and clean up.

How do I install it?

Use the install box at the top. First install Snowflake CLI and add a connection with snow connection add. The box has Claude Code, Codex and the skills CLI. It has no Claude app upload, because the workflow runs the snow command on your machine.

Does it need a Snowflake account?

Yes. It validates DDL by sending it to Snowflake with snow sql on a connection you have configured.

Will it change my data?

It creates a temporary semantic view and then creates or alters the real one. SKILL.md does not describe changes to table data. Check the DDL before it is applied.

Is this an official Snowflake skill?

It is a community skill in GitHub's awesome-copilot repo, not in a Snowflake repo. Snowflake's own MCP server has a separate page.

Does it invent synonyms and comments?

Not without approval. SKILL.md says to read existing Snowflake comments first and to ask before creating or drafting any.

Is it safe?

It runs SQL on your Snowflake account, so use a development role first. It creates and alters a semantic view, and SKILL.md has no approval step before the final apply. See the permissions section.

Information

updated Nov 2, 2026

Repository: github/awesome-copilot
Folder: skills/snowflake-semanticview

github/awesome-copilot skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 1

snowflake-semanticview SKILL.md · github.com