snowflake-query-credit-guardrail

Offline Snowflake credit guardrail for query-history exports. It converts warehouse burn, cache-miss repeats, oversized warehouses, owner gaps, and untagged spend into a board-readable remediation queue.

Why this exists

Snowflake cost work often collapses into monthly totals after the damage is done. Leadership needs the earlier view: which queries are wasting credits, which warehouse lanes are over-provisioned, which teams need tagging hygiene, and which fixes should move before the next finance review.

What it shows

• Query-level credit exposure and avoidable-credit estimate.
• Warehouse and owner lanes ranked by waste pressure.
• Remediation actions for cache reuse, warehouse sizing, tagging, and owner routing.
• SQL extraction template that keeps the repo credential-free.
• Static proof page for GitHub Pages and portfolio indexing.

What this product does

Snowflake Query Credit Guardrail turns query-history exports into a finance-readable guardrail: which queries are burning credits, which warehouses are over-provisioned, which owners need tagging hygiene, and what should be remediated before the next spend review.

For a SaaS go-to-market analyst, the product frames data-platform spend as an operating signal. Reporting teams cannot protect revenue narratives if warehouse-backed dashboards are slow, expensive, or impossible to explain. For a SaaS value architect, the value is a repeatable cost-governance motion: recover avoidable credits, protect board reporting, assign owners, and convert raw query history into a credible savings narrative.

Technically, this repo ships a credential-free CLI, JSON and markdown output, SQL extraction template, deterministic fixture analysis, unit tests, static site generation, and a CI safety scan for local usernames and deploy secrets. It follows the broader Kinetic Gain pattern: turn hidden operational drag into named lanes, evidence, owner accountability, and a board-readable next action.

Local run

python -m pip install -e .
python -m unittest discover -s tests
snowflake-query-credit-guardrail fixtures/snowflake-query-credit-sample.json
python scripts/build_site.py

CLI

snowflake-query-credit-guardrail fixtures/snowflake-query-credit-sample.json --format markdown
snowflake-query-credit-guardrail fixtures/snowflake-query-credit-sample.json --format json

Data contract

The CLI accepts JSON shaped as either an array or { "queries": [...] }. CSV input is also accepted when headers match the field names below.

Required or useful fields:

• query_hash
• warehouse
• warehouse_size
• owner
• business_unit
• credits
• bytes_scanned
• execution_seconds
• rows_produced
• cache_hit
• tag_status
• classification

See sql/query_credit_guardrail.sql for an account-usage extraction pattern.

Kinetic Gain fit

This is a platform-and-company signal repo for Snowflake, FinOps, data engineering, and executive operating reviews. It complements the GCP, BigQuery, Power BI, Tableau, and broader platform-governance surfaces in the Kinetic Gain portfolio.