Snowflake Snowpark Dbt

What it does

When your agent uses it

• Writing server-side Snowpark pipelines
• Registering UDFs
• Stored procedures
• Choosing dbt materializations

Example prompts

• “/snowflake-snowpark-dbt”

Requirements

• Python 3

Workflow steps

8 steps, taken from the first numbered list in SKILL.md.

1. 1Snowpark: open a session — Build a Session from environment-scoped credentials, specifying role, warehouse, database, and schema explicitly.
2. 2Snowpark: express transforms with the DataFrame API — Prefer .filter(), .select(), .group_by().agg(), and .join() over raw SQL strings for…
3. 3Snowpark: push compute server-side — Use scalar UDFs for row-wise logic, vectorized (pandas) UDFs for ML inference, UDTFs when one input…
4. 4dbt: model in layers — Staging models (stg_*) rename and type-cast; mart models express business logic on top of staging.
5. 5dbt: choose a materialization — view for cheap logic, table only when reads are frequent, incremental for large fact tables, dynamic_table…
6. 6dbt: define sources and tests — Declare sources in _sources.yml with freshness thresholds; add unique/not_null tests on key columns.
7. 7dbt: run selectively — Use dbt run --select model+ (model and downstream) or +model (model and upstream) instead of full-project runs…
8. 8dbt: build and validate — Run dbt build (run + test in dependency order) before merging, and dbt docs generate to keep documentation…

What it can do on your machine

Read from SKILL.md and the folder at commit 7682ca7. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • dbt
  • pip

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use pip, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names these keys or tokens, usually read from environment variables:

  • SNOWFLAKE_PASSWORD

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Snowflake Snowpark Dbt loads about 2.5k tokens when it runs. Until then it costs about 90 tokens; SKILL.md has 753 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~90

When it runs · the whole SKILL.md, loaded when a task matches

~2.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from Mindrally/skills at commit 7682ca7, republished under its Apache-2.0 licence (© Mindrally). 753 words, ~2,472 tokens.

name

snowflake-snowpark-dbt

description

Best practices for Snowpark Python (DataFrames, UDFs, UDTFs, stored procedures) and dbt with the dbt-snowflake adapter. Use when writing server-side Snowpark pipelines, registering UDFs or stored procedures, choosing dbt materializations, configuring incremental models, or setting up sources and tests for a Snowflake-backed dbt project.

metadata.maintainer

Mindrally

metadata.source

https://github.com/Mindrally/skills

import os
from snowflake.snowpark import Session

session = Session.builder.configs({
    "account": os.environ["SNOWFLAKE_ACCOUNT"],
    "user": os.environ["SNOWFLAKE_USER"],
    "password": os.environ["SNOWFLAKE_PASSWORD"],
    "role": "my_role", "warehouse": "my_wh", "database": "my_db", "schema": "my_schema",
}).create()

df = session.table("customers")
df_filtered = df.filter(df["region"] == "US").select("name", "email", "revenue")
df_agg = df.group_by("region").agg(sum("revenue").alias("total_revenue"))
df_agg.show()

from snowflake.snowpark.functions import udf

@udf(name="normalize_email", replace=True)
def normalize_email(email: str) -> str:
    return email.strip().lower() if email else None

import pandas as pd
from snowflake.snowpark.functions import udf

@udf(name="predict_score", packages=["scikit-learn", "pandas"], replace=True)
def predict_score(features: pd.Series) -> pd.Series:
    import pickle, sys
    model = pickle.load(open(sys.path[0] + "/model.pkl", "rb"))
    return pd.Series(model.predict(features.values.reshape(-1, 1)))

from snowflake.snowpark.types import StructType, StructField, StringType

class Tokenizer:
    def process(self, text: str):
        for token in text.split():
            yield (token,)

tokenize = session.udtf.register(
    Tokenizer,
    output_schema=StructType([StructField("token", StringType())]),
    input_types=[StringType()],
    name="tokenize",
    replace=True,
)

from snowflake.snowpark import Session
from snowflake.snowpark.functions import sproc

@sproc(name="daily_etl", replace=True, packages=["snowflake-snowpark-python"])
def daily_etl(session: Session) -> str:
    raw = session.table("raw_events")
    cleaned = raw.filter(raw["event_type"].is_not_null())
    cleaned.write.mode("overwrite").save_as_table("cleaned_events")
    return f"Processed {cleaned.count()} rows"

my_project:
  target: dev
  outputs:
    dev:
      type: snowflake
      account: myaccount
      user: myuser
      password: "{{ env_var('SNOWFLAKE_PASSWORD') }}"
      role: transformer
      database: analytics
      warehouse: transforming
      schema: public
      threads: 4

{{ config(materialized='dynamic_table', snowflake_warehouse='transforming', target_lag='1 hour') }}
SELECT customer_id, SUM(amount) AS lifetime_value FROM {{ ref('stg_orders') }} GROUP BY 1

{{
  config(
    materialized='incremental',
    unique_key='event_id',
    incremental_strategy='merge',
    on_schema_change='sync_all_columns'
  )
}}
SELECT * FROM {{ ref('stg_events') }}
{% if is_incremental() %}
  WHERE event_timestamp > (SELECT MAX(event_timestamp) FROM {{ this }})
{% endif %}

sources:
  - name: raw
    database: raw_db
    schema: jaffle_shop
    tables:
      - name: customers
        loaded_at_field: _loaded_at
        freshness:
          warn_after: {count: 12, period: hour}
          error_after: {count: 24, period: hour}

models:
  - name: stg_customers
    columns:
      - name: customer_id
        tests: [unique, not_null]

{% macro generate_schema_name(custom_schema_name, node) %}
  {% if custom_schema_name %}{{ custom_schema_name | trim }}{% else %}{{ target.schema }}{% endif %}
{% endmacro %}

© Mindrally, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in snowflake-snowpark-dbt of Mindrally/skills.

Open the folder on GitHubat commit 7682ca7