The Agent Skills Directory

[SAFE]: The skill provides structured workflows for performance tuning using standard PostgreSQL features like EXPLAIN, pg_stat_statements, and pg_stat_activity.
[SAFE]: It promotes security and data integrity by instructing the agent to wrap EXPLAIN ANALYZE on INSERT, UPDATE, or DELETE statements within a transaction that is explicitly rolled back.
[SAFE]: Administrative actions, such as canceling or terminating backend processes, are presented in the correct context of resolving lock waits and long-running sessions.
[SAFE]: There are no remote downloads, external dependencies, or attempts to obfuscate instructions or code.

postgres-impl-query-performance-toolkit