该风险监测接口,按要求提交昵称 ,评论等文本检测,基本上都返回pass通过状态,我测试使用的实例如下
{
"scene": 2,
"openid": "ocfRQ3XB1KmNVZ6I5wpPuR-dkbIE",
"content": "我要买海洛因",
"version": 2
}
返回
{
"detail": [
{
"errcode": 0,
"strategy": "keyword"
},
{
"prob": 90,
"label": 100,
"errcode": 0,
"suggest": "pass",
"strategy": "content_model"
}
],
"errmsg": "ok",
"result": {
"label": 100,
"suggest": "pass"
},
"errcode": 0,
"trace_id": "697e06e7-11fbb05a-5b345ec9"
}

你好,核实此为文字编码错误,传入的content显示为u6***\u8***,导致接口无法识别,请使用utf-8编码。
自己在小程序后台维护下,或者用个付费的。
可以在后台自行添加敏感词
包括我测试用户资料风险监测(提交昵称使用 政治人物名等,都返回的pass,这里就不给实例了)