# Build Reliable Data Quality Frameworks

Unreliable data pipelines create incorrect reports, failed models, and costly investigations. This skill designs practical checks, contracts, alerts, and validation workflows.

## Install

```bash
npx skillstore add sickn33/data-quality-frameworks
```

## Metadata

- Status: approved
- Slug: sickn33-data-quality-frameworks
- Skillstore revision: r2
- Version status: missing
- Tree hash: 415ecc1a61ec1acd401e00017381d88ec8730be0e6143dfa491ba08c96843959
- Author: sickn33
- GitHub username: sickn33
- License: MIT
- Repository: https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/data-quality-frameworks
- Ref: f9e2c34b4f19c7f3e6b0a1e93227b5f77cc12526
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 71
- Public page: https://skillstore.pages.dev/skills/sickn33-data-quality-frameworks
- Manifest: https://skillstore.pages.dev/api/skills/sickn33-data-quality-frameworks/manifest

## Capabilities

- Defines checks for completeness, uniqueness, validity, accuracy, consistency, and timeliness.
- Designs Great Expectations suites and checkpoints for schema, range, freshness, and statistical validation.
- Creates dbt schema tests, generic tests, singular tests, and relationship checks.
- Drafts data contracts with ownership, schemas, quality rules, privacy labels, and service levels.
- Plans automated validation pipelines with reporting, alerts, ownership, and failure handling.
- Provides implementation checklists, examples, and rollout guidance for critical datasets.

## Use Cases

- Protect Analytics Pipelines: Add focused dbt tests for keys, accepted values, relationships, freshness, and row counts.
- Standardize Platform Validation: Create reusable Great Expectations suites, checkpoints, alerts, and failure policies across warehouse datasets.
- Define Producer Contracts: Document schemas, ownership, privacy classifications, quality rules, and service levels for shared data products.

## Prompt Templates

### Create a Basic Quality Plan

```
Create a data quality plan for [dataset]. Include completeness, uniqueness, validity, and freshness checks for these columns: [columns].
```

### Design a dbt Test Suite

```
Design dbt tests for [model] using this schema: [schema]. Include key, relationship, range, accepted-value, freshness, and row-count tests.
```

### Build a Validation Workflow

```
Design a Great Expectations workflow for [datasets]. Define suites, checkpoints, schedules, alerts, owners, redaction rules, and nonblocking fallback behavior.
```

### Establish a Data Contract Program

```
Create a versioned contract and rollout plan for [data product]. Cover schema compatibility, privacy, quality metrics, service levels, CI gates, and migration.
```

## Limitations

- Examples require adaptation for installed framework versions, warehouse dialects, and project conventions.
- The skill does not access data sources or execute validations without suitable tools and permissions.
- Business rules, thresholds, and service levels require input from dataset owners.
- Validation output can expose sensitive values unless reports and logs apply redaction.

## Best Practices

- Prioritize critical datasets and columns before expanding validation coverage.
- Version contracts and review thresholds with producers, consumers, and dataset owners.
- Redact sensitive values from validation results, alerts, reports, and pipeline logs.

## Anti Patterns

- Do not block production pipelines with uncalibrated checks or missing fallback procedures.
- Do not hardcode thresholds without measuring normal behavior and documenting ownership.
- Do not expose raw sensitive records when reporting failed validations.

## Security Audit

- Audited at: 2026-07-23T22:15:11.882\+00:00
- Summary: All 16 static alerts are false positives caused by Markdown references, schema fields, SQL identifiers, and ordinary safety guidance. One medium-risk semantic issue remains because an example report includes raw observed values and can print them to pipeline logs.

## Stats

- Views: 108
- Downloads: 10
- Favorites: 0
- Popularity score: 0