The Agent Skills Directory

[SAFE]: The skill is entirely focused on engineering best practices for avoiding production outages during system evolution. It implements a rigorous safety framework ('Domain Law') requiring state definitions and verification gates before any changes occur.
[SAFE]: No prompt injection or behavior override patterns were detected. The instructions are professional and aligned with the stated purpose of safe engineering guidance.
[SAFE]: No data exfiltration or credential theft attempts were found. The skill references sensitive concepts like database transaction logs (WAL, binlog) and CDC pipelines purely in an architectural and operational safety context.
[SAFE]: All external URLs point to official, well-known documentation for reputable technologies (PostgreSQL, MongoDB, Debezium, Kubernetes, IETF, Apache Kafka, Elasticsearch, etc.) or academic papers (ArXiv). No suspicious or obfuscated links were identified.
[SAFE]: There are no remote code execution patterns, unauthorized package installations, or dynamic code execution blocks. The worked example correctly illustrates a safe column rename using standard expand-and-contract sequencing without dangerous side effects.

migration-evolution