linkfox-temu-returns-refunds-eu

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill shows and encourages embedding secrets verbatim in shell/CLI commands (exporting LINKFOXAGENT_API_KEY and passing "accessToken" inline in the JSON argument to python), which forces an LLM or user to place secrets directly into generated output/commands — an exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.78). At runtime the required workflow calls Temu/LinkFox gateway endpoints (POST /temu/proxy and parses response.body via JSON.parse(body)), then prints/summarizes that gateway-returned JSON (which can include outsider-authored free text like errorMsg, afterSalesReasonDesc, buyerComment, address fields) into the agent’s LLM context via stdout/inline output.

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

• Direct money access detected (high risk: 1.00). 该 skill 明确面向「退货与退款（Returns & Refunds）」场景，描述已接入多条 Partner Return and Refund 接口并内置了对 Temu Partner API 的网关（POST https://tool-gateway.linkfox.com/temu/proxy 与文件下载等）。这些接口/脚本集合专门用于售后退货与退款相关操作（即直接涉及发起/管理退款的业务能力），因此属于明确面向执行财务/退款操作的工具链，具有直接财务执行的能力。