quality-gates
Use when establishing tests, monitoring, and incident response for analytics models.
When & Why to Use This Skill
The Analytics Quality Gates skill provides a comprehensive framework for ensuring the reliability and integrity of data models and dashboards. By integrating automated testing, real-time alerting, and structured incident response, it helps data teams transition toward a 'DataOps' approach, minimizing downtime and preventing 'silent failures' in business-critical analytics pipelines.
Use Cases
- Automated Data Quality Monitoring: Establishing validation steps for schema consistency, data freshness, and volume thresholds to ensure dashboards reflect accurate information.
- Incident Response & Playbooks: Utilizing structured triage checklists and communication templates to rapidly resolve data breaks and maintain stakeholder trust.
- CI/CD Integration for Data: Implementing quality gates within deployment pipelines (e.g., dbt Cloud or GitHub Actions) to block faulty models or code from reaching production.
- Post-Incident Analysis: Conducting root cause analysis (RCA) and documenting remediation tasks to build a long-term knowledge base and prevent recurring data issues.
| name | quality-gates |
|---|---|
| description | Use when establishing tests, monitoring, and incident response for analytics |
Analytics Quality Gates Skill
When to Use
- Designing validation steps for new models or dashboards.
- Setting up automated data quality monitoring.
- Running incident reviews after data breaks.
Framework
- Test Coverage – schema, freshness, unique, referential, accepted values, volume thresholds.
- Alerting – severity tiers, alert channels, on-call rotation, escalation policies.
- Incident Response – triage checklist, communication templates, resolution targets.
- Change Management – approval workflow, rollback plan, audit logging.
- Postmortems – root cause analysis, remediation tasks, knowledge base updates.
Templates
Templates
- QA Checklist: See
assets/qa_checklist.mdfor validation steps. - Quality matrix (model/table → tests → owner → SLA).
- Incident playbook (trigger, response steps, communication log).
- Change request form with risk assessment.
Tips
- Tie quality gates to CI/CD or dbt Cloud jobs to block bad deploys.
- Keep alert fatigue low by tuning thresholds.
- Document every incident for future prevention.