The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute numerous shell commands using tools such as nmap, hydra, medusa, nc, and msfconsole for the purpose of network auditing.
[PRIVILEGE_ESCALATION]: The skill instructs the agent to use sudo to perform specific nmap scans that require administrative privileges.
Evidence: sudo nmap -sV -sC -p 21 <IP> in SKILL.md.
[EXTERNAL_DOWNLOADS]: The skill uses wget to recursively download files and directories from remote FTP servers that are being audited.
Evidence: wget -m ftp://anonymous:anonymous@<IP> and related commands in SKILL.md and references/ftp-commands.md.
[INDIRECT_PROMPT_INJECTION]: The process of enumerating and downloading content from external FTP servers creates a surface for indirect prompt injection, as the agent ingests potentially attacker-controlled data.
Ingestion points: File names, directory listings, and file contents retrieved during enumeration and recursive downloads.
Capability inventory: The agent is capable of executing network operations and system commands.
Boundary markers: No specific delimiters or instructions to treat external data as untrusted are included in the skill.
Sanitization: No sanitization or validation logic is defined for data processed from remote FTP targets.

ftp-pentesting