imap-pentesting

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill prompt repeatedly shows and instructs embedding plaintext credentials into commands (e.g., "a1 LOGIN username password" and curl --user user:pass), which would require the model to include secret values verbatim in generated outputs and thus enables secret exfiltration.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document provides explicit, actionable instructions for unauthorized access, credential theft, and data exfiltration from IMAP servers (brute-forcing, NTLM info extraction, automated download of mail/attachments, and exploitation of RCE CVEs), indicating clear malicious/abusive intent or facilitation thereof.

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

• Third-party content exposure detected (medium risk: 0.30). Skill’s required runtime workflow in Phase 4 performs IMAP mailbox listing/search and then FETCH downloads and parses message bodies/headers, meaning outsider-authored email text from an external server the user connects to is ingested as LLM-readable content at runtime.