ssh-penetration-testing

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill includes hardcoded/example passwords and key blobs and contains code that prints username:password pairs and echoing key material, which would require an LLM to reproduce secret values verbatim in outputs/commands (high exfiltration risk).

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). These URLs point to web-accessible SSH private keys and a backup archive on target.com (exposed credentials/data), which is a high-risk indicator of sensitive data leakage that attackers could use to gain unauthorized access.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document contains explicit, actionable instructions for unauthorized access techniques (brute-force, credential harvesting, reverse tunnels) and persistence (adding keys), representing high-risk malicious capabilities despite an "authorized use" disclaimer.

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

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly instructs post-exploitation changes (echoing an SSH key into authorized_keys), editing system configuration (/etc/proxychains.conf), and running privileged checks (sudo -l), which direct the agent to modify and persist state on a machine.