linkfox-temu-ads-eu

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

• Third-party content exposure detected (low risk: 0.10). 该 skill 的运行时从用户提供的 JSON 参数中提取 type/params并直接调用网关 POST https://tool-gateway.linkfox.com/temu/proxy（scripts/temu_eu_proxy.py→_eu_ads_script.run_cli→_temu_eu_common.eu_proxy_call→call_temu_api），但这属于对方只把内容当作结构化请求负载转发而非读取外部免费文本来源；LLM 实际也仅打印网关返回的 JSON（由 Temu/网关生成的字段/错误信息），不对外部“被投毒的文本源”做主动检索/抓取。

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I scanned the skill files for literal, high-entropy credentials. Most token examples are placeholders (e.g. "", "TOKEN", "PASTE_TOKEN") or masked examples and are intentionally ignored per the rules. However I found a literal base64-like string assigned to _LOGIN_FIXED_UID in scripts/onboarding.py. It's a non-trivial, high-entropy constant (not a placeholder) used as a fixed uid header when no access token is present ("实测可用" in code comment). Because it is a literal token-like value that may be usable in requests (i.e., could provide access/identify a client), I flag it as a potential secret.

No other high-entropy, direct credentials (API keys, private keys, long tokens) are present — only placeholders and example tokens which are ignored per instructions.