web-enumeration

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill fetches .env and log files and explicitly extracts and prints environment variables, API keys, JWTs and other secret values verbatim (e.g., print KEY {m.strip()}, print DONE {r.text[:150]}, curl ... | head), which requires the agent/LLM to handle and output secrets directly.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.85). These URLs include an obviously malicious domain plus multiple endpoints that expose sensitive files (.env, .git, laravel logs) which are high-risk because they can leak credentials or be used to stage further compromise/malware delivery.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This content is an offensive reconnaissance guide that explicitly instructs locating and exfiltrating sensitive files and credentials, bypassing protections, and harvesting tokens/JWTs — high risk for malicious abuse.

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

• Third-party content exposure detected (medium risk: 0.30). In SKILL.md, the runtime workflow uses requests.get to fetch outsider-controlled HTTP responses from arbitrary target.com-constructed paths (e.g., /.env, /.git/config, and /storage/logs/laravel.log) and then ingests free-text from r.text/.text (HTML body samples and full logs) for keyword matching and printing.