The Agent Skills Directory

[SAFE]: The skill provides standard PostgreSQL monitoring functionality using read-only SQL queries to fetch performance metrics from system tables (e.g., pg_stat_activity, pg_stat_statements).
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze database metrics. While this represents a data ingestion surface, the risk is negligible as it uses specific monitoring queries and does not execute arbitrary code based on the data retrieved.
[REMOTE_CODE_EXECUTION]: The provided validation script (scripts/validate.py) performs static checks on file structure and YAML configuration using safe loading practices. It does not perform network operations or execute external code.

postgresql-monitoring