dbt-labs/dbt-agent-skills

A curated collection of Agent Skills for working with dbt, to help AI agents understand and execute dbt workflows more effectively.

Quality

82%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Risky

Do not use without reviewing

Writing project documentation

Name: dbt-labs/dbt-agent-skills
Rating: 65.91 (1 reviews)
Author: dbt-labs

Never generate documentation which simply restates the entity's name. Describe why, not just what.

Inspect the data before writing documentation about it, using discovering-data.

Table level

Describe the grain of the table, its purpose and any edge cases

Bad:

models:
  - name: active_customers
    description: All customers who are active

Good:

models:
  - name: active_customers
    description: The `customers` table pre-filtered for easier analytics. One row per customer whose contract_expiry_date is null or in the future

Column level

Calculated fields should include a brief description of the transformation and its purpose.

Bad:

models: 
  - name: customers
    columns: 
      - name: customer_id
        description: The customer's identification number

Good:

models: 
  - name: customers
    columns: 
      - name: customer_id
        description: Users older than 2020-02-16 have `v1_` prefixed to their customer ID due to the platform migration.