wechat-search

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

• Third-party content exposure detected (medium risk: 0.30). The workflow queries WeChat Official Account articles via the RedFox API, ingesting third-party article titles, summaries, and content from public posts returned in search results.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill makes runtime HTTP requests to redfox.hk and recommends installation of npm package @alipay/agent-payment, which lacks verified first-party publisher provenance and constitutes an external operational dependency.

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

• Direct money access detected (high risk: 0.70). The skill documents a WeChat search tool that retrieves trending articles via an API. It mentions handling low-balance situations by triggering an Alipay recharge workflow (via alipay-bot trigger-payment-signal and submit-payment) with cashier URLs and order amounts. This constitutes a direct payment/recharge operation. Although instructions mention polling payment status and handling signals, no independent, service-enforced authentication or authorization mechanism is documented specifically for the payment/recharge route itself (the API key is for the search API, and Alipay payment routing lacks documented independent service-side checks beyond bot instructions). Therefore, it qualifies as a direct financial operation without established qualifying independent control.