smtp-pentesting

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.80). The GitHub repository (https://github.com/serain/mailspoof) is potentially suspicious because it is a third‑party tool for SPF/DMARC spoofing from an individual account (could distribute offensive scripts/binaries), while the other two URLs are benign operational references (an MTA-STS policy file and a well-known SPF validator).

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This is an explicit offensive SMTP attack playbook: it contains step-by-step instructions and code to enumerate users, steal credentials, brute-force/collect NTLM hashes, craft/send spoofed/phishing mail (including DKIM forgery), use CRLF/SMTP smuggling to inject recipients/headers, exploit known MTA RCEs, and use SMTP as a data exfiltration channel — all high-risk malicious behaviors.

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md 在 Phase 1/2/3/4/5/6/7 通过 Nmap/telnet/openssl/dig 等对外部目标的 SMTP 服务进行探测与交互（例如读取 Banner、EHLO 扩展、VRFY/EXPN/RCPT 响应、以及邮件头/退信等），这些读取内容来自“被攻击方/外部服务”而非可信的 first-party提供的内容，且外部方可被外人以自由文本形式影响（如在 SMTP/邮件协议返回或 Web 表单注入的邮件头/内容中）。

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.70). The skill explicitly directs post‑exploit inspection of system configuration files (e.g. /etc/postfix/*), references using root‑level tools like tcpdump, and highlights master.cf script execution paths—actions that involve accessing or exploiting system files and can change or compromise the host state.