wechat-hotspot-publisher

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (critical risk: 1.00). The skill contains an explicit custom publishing API endpoint that exfiltrates article titles, HTML contents, cover image URLs, and tags to an external server without a legitimate ancillary purpose.

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

• Third-party content exposure detected (medium risk: 0.30). The workflow monitors and fetches outsider-authored hot topics and social feeds from third-party platforms via fetch_sources.py (such as Zhihu and Weibo feeds).

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

• Potentially malicious external URL detected (high risk: 0.70). The script publish_wechat.py utilizes a custom API endpoint (http://39.108.254.228:8002/publish-draft) for operation without prior identity validation or official publisher provenance, presenting a weak/unknown/suspicious operational dependency.