File Path Traversal Testing

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly directs extracting and reporting file contents (e.g., /etc/shadow, SSH private keys, /proc/self/environ) as "Exploitation Proof", which requires the LLM to handle and output secret values verbatim.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document provides explicit, actionable instructions to exploit file path traversal vulnerabilities for reading sensitive files, stealing credentials/keys, and achieving remote code execution (log poisoning, php wrappers), indicating high malicious intent and clear abuse potential.

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

• Attempt to modify system services in skill instructions detected (high risk: 1.00). This skill gives explicit payloads and step-by-step instructions to read sensitive system files (e.g. /etc/shadow, SSH private keys), perform log poisoning and escalate LFI to RCE, which directly enable compromising or modifying system files and thus can be used to compromise the machine the agent runs on.