xiaohongshu-prohibited-word

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (critical risk: 1.00). The skill includes automated payment triggers and polling via an external tool (alipay-bot) linked to a third-party gateway, which constitutes a financial transaction flow inside a non-financial content-checking skill.

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

• Third-party content exposure detected (low risk: 0.10). The required workflow ingests external web content via URL scraping (extract_from_web) and local document parsing (extract_from_file), making it a low-risk vector for indirect prompt injection from fetched pages or files.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill makes an operational runtime HTTPS POST request to an external domain (redfox.hk) governed by third-party infrastructure and configuration, representing an unknown/unverified third-party service dependency with high risk.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 0.70). The skill performs a direct financial operation by handling automated payment execution (via Alipay recharge signal triggers and payment completion polling) when API credit is insufficient. No qualifying independent access control is documented for this direct payment route.