remote-access-enumeration

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.70). The prompt explicitly claims "enumeration only" but includes directives that perform or encourage exploitation/brute-force (e.g., nmap vnc-brute, vsftpd backdoor checks and shell triggers), which are instructions outside the skill's stated scope and thus deceptive/inconsistent.

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt instructs the agent to capture discovered credentials and include them verbatim (e.g., add_credential(username=..., secret=...) and reporting findings in the return summary), which requires the LLM to output secret values directly.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill contains explicit, actionable instructions for credential harvesting and exploiting pre-auth/remote-code-execution/backdoor vectors (e.g., vsftpd backdoor, ProFTPD mod_copy, BlueKeep, regreSSHion, VNC no-auth) and will enable unauthorized data exfiltration and remote compromise if misused.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow ingests outsider-authored free text when it runs nmap NSE scripts (e.g., ftp-anon, ftp-syst, ssh-hostkey, rdp-ntlm-info, vnc-info) against targets, because the resulting human-readable service banners/script outputs are derived from remote third-party systems’ text and are fed into the agent context for parsing/reporting.