Snowflake Native App · Beta

SQL Branch Coverage
for your dbt Models

Measure C1 (branch) coverage of dbt SQL models — directly inside Snowflake.
CI/CD ready via stored procedure. Zero infrastructure.

Coming Soon on Marketplace

snowsql (CI/CD)

# Call the stored procedure from your CI/CD pipeline
snowsql -q "CALL code.analyze_coverage('TARGET_DB', 'dbt_schema', '{}')"

[OK]  fact_sales              100.0%  (5/5 branches)
[OK]  dim_customers            92.3%  (12/13 branches)
[FAIL] customer_segmentation    66.7%  (2/3 branches)
       └─ L12: segment = 'New'  ← uncovered branch

[OK]  Overall: 93.75% coverage

What is SQL Branch Coverage?

Every condition has two outcomes.
Are you testing both?

customer_segmentation.sql

SELECT
  customer_id,
  CASE
    WHEN total_purchases > 10  THEN 'VIP'     ✓ True
    WHEN total_purchases >  3  THEN 'Regular' ✓ True
    ELSE                           'New'      ✗ never reached
  END AS segment
FROM customers
WHERE status = 'active'
  AND created_at >= '2023-01-01'
  -- ↑ WHERE clause as a whole: True ✓ | False ✗

60%

Branch Coverage

Uncovered Branches

C1 Coverage rule:

Each condition (CASE WHEN, WHERE clause) must be evaluated as both True and False. AND/OR sub-conditions are not tracked individually — the full WHERE expression is one branch.

Features

Everything you need for
data pipeline quality

🔬

C1 Branch Coverage

Analyzes CASE WHEN, WHERE, and boolean conditions in dbt SQL. Measures True/False coverage per expression — not per sub-condition.

🏔️

Native Snowflake App

Runs entirely inside your Snowflake account. No external services, no data leaves your environment. Deployed via Snowflake Marketplace.

🚀

CI/CD Ready

Integrate coverage checks into any CI/CD pipeline by calling the Snowflake stored procedure. Returns structured JSON results for downstream processing.

📊

Multiple Report Formats

JSON (machine-readable), JUnit XML (CI integration), Markdown (PR comments), and HTML (visual browser report).

🖥️

Streamlit UI

Built-in Streamlit dashboard for interactive analysis, run history, and HTML report download.

🌐

i18n Support

Full Japanese and English support across UI, error messages, and report output.

🔧

Exclusion Rules

Fine-grained control with .dbt-coverage.yml to exclude models, schemas, or specific conditions from analysis.

How it works

Three steps to coverage insights

Upload dbt Artifacts

Run dbt compile and upload the target/ directory as target.zip to your Snowflake Stage.

dbt compile
Compress-Archive -Path ./target -DestinationPath target.zip
snowsql ... PUT file://target.zip @dbt_coverage_app.code.dbt_artifacts

Analyze via Stored Procedure

Call the analyze_coverage stored procedure — from CI/CD script or the Streamlit UI.

CALL code.analyze_coverage(
  '@dbt_coverage_app.code.dbt_artifacts/target.zip',
  '*',     -- model selector
  'api',   -- caller
  'en'     -- language
);

Get Structured Results

Receive a detailed JSON report. Export as JUnit XML for CI, Markdown for PR comments, or HTML for visual review.

{
  "status": "success",
  "summary": {
    "coverage_rate_percent": 93.75,
    "models_passed": 5,
    "models_failed": 1
  },
  "models": [ ... ]
}

CI/CD Integration

Integrate into CI/CD via
Snowflake Stored Procedure

name: dbt Coverage Analysis
on: [push, pull_request]
jobs:
  coverage:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Run coverage analysis
        run: |
          snowsql -q "CALL code.analyze_coverage( \
            'TARGET_DB', 'dbt_schema', '{}')" \
            -o output_format=json
      - uses: actions/upload-artifact@v4
        with:
          name: coverage-reports
          path: coverage_reports/

# Call the stored procedure directly with snowsql
snowsql -q "CALL code.analyze_coverage('TARGET_DB', 'dbt_schema', '{}')" `
  -o output_format=json

# Parse the JSON result and use exit code on your CI/CD side
# 0 = success, non-zero = error

✅ Call via snowsql from any CI/CD environment
✅ Returns structured JSON for downstream processing
✅ Markdown summary to GitHub Step Summary
✅ HTML report as build artifact
✅ Sample repository with GitHub Actions integration — coming soon

Stored Procedure API

Simple SQL interface

All functionality is exposed as Snowflake Stored Procedures — callable from any SQL client, BI tool, or script.

code.analyze_coverage

Run branch coverage analysis on your dbt artifacts.

CALL code.analyze_coverage(
  stage_path,      -- '@.../target.zip'
  model_selector,  -- '*' or 'fact_sales'
  caller,          -- 'api' for CI/CD
  lang             -- 'en' or 'ja'
);

code.list_history

Retrieve recent analysis run history.

CALL code.list_history(20);
-- Returns last 20 runs with
-- coverage rates, timestamps,
-- model counts, and status.

code.generate_html_report

Generate an HTML report from analysis results.

CALL code.generate_html_report(
  result_json,  -- from analyze_coverage
  lang          -- 'en' or 'ja'
);
-- Returns full HTML string

SQL Branch Coverage for your dbt Models

Every condition has two outcomes.Are you testing both?

Everything you need fordata pipeline quality