quality-control

Production verification quality monitoring uses LLM-as-Judge evaluation to assess verification results through adaptive sampling and confidence-based triage. This approach maintains verification standards by optimizing review load while preserving confidence in the assessed outputs.

Drift detection in production verification identifies when monitoring results deviate from expected patterns or encounter novel failure modes. You need drift detection when maintaining oversight of production systems to trigger maintenance or evolution loops automatically when issues are detected.

Setting up adaptive sampling rules for production verification involves configuring the QC skill in your workspace to monitor production results. The skill applies adaptive sampling strategies that optimize review load while preserving confidence in the verification outcomes.

Yes, you can trigger evolution loops automatically when monitoring detects drift, failures, or novel patterns in verification results. The QC skill supports triggering evolution to address these issues, enabling automated maintenance and continuous improvement of production verification workflows.

The quality-control skill includes dashboards and auditing features to support governance and incident response. These capabilities provide visibility into production verification monitoring, drift detection, and the evolution loop triggers activated during oversight.