database-enumeration
Pass
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill leverages standard network and database clients (e.g., nmap, mysql, psql, mongosh, redis-cli) alongside specialized penetration testing scripts like mssqlclient.py and odat to conduct service enumeration.
- [REMOTE_CODE_EXECUTION]: The skill contains logic to verify and execute commands on target systems through database-specific features like xp_cmdshell, User-Defined Functions (UDF), and COPY PROGRAM. It also includes routines to achieve persistent access on Redis targets by writing webshells or SSH authorized keys.
- [DATA_EXFILTRATION]: It retrieves and logs database contents (e.g., user tables, configurations) by redirecting query output to local temporary files and then reading them for analysis.
- [PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection when processing untrusted database output.
- Ingestion points: Data retrieved from database queries and read from temporary files (SKILL.md).
- Boundary markers: Not present.
- Capability inventory: Extensive command execution capabilities including file system access and networking.
- Sanitization: No sanitization is performed on data retrieved from external database services.
Audit Metadata