web-region-audience-analyzer

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly shows passing an NLP Cloud token as a command-line argument (--nlpcloud-token YOUR_TOKEN), which instructs embedding a secret verbatim in a command and therefore risks exfiltration.

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches and crawls arbitrary public websites (see SKILL.md "Run the provided Python script on a target URL" and scripts/analyze_webpage.py functions fetch_html / crawl_site), ingests the page text (extract_signals) and even sends it to an AI analyzer (analyze_content_with_ai), and uses those results to compute region/audience and actionable recommendations, so untrusted third‑party page content can materially influence analysis and downstream actions—enabling indirect prompt injection.