sqlserver-monitoring
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues were identified in the SQL scripts or instructions provided. The skill utilizes standard system views and stored procedures for database administration and monitoring.
- [PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection by retrieving database logs and query text that may contain external input.
- Ingestion points: Database error logs (xp_readerrorlog) and active SQL query text (sys.dm_exec_sql_text) from system views in SKILL.md.
- Boundary markers: No specific delimiters are used to wrap external content.
- Capability inventory: The skill is limited to diagnostic SQL queries and does not have file-system, shell, or network access.
- Sanitization: No sanitization is applied to the retrieved text data.
Audit Metadata