OpenAI 获取项目安全警报

获取项目安全警报

get /safety/alerts/{id}

获取属于当前通过身份验证的 API 项目的安全警报。

路径参数

• id: string

项目安全警报 ID。

maxLength: 38

返回值

• SafetyAlert object { id, created_at, error_type, 7 more }

• id: string

• created_at: integer

format: unixtime

• error_type: "potentially_unintended_data_transfer" or "potentially_unintended_data_access" or "potentially_unintended_destructive_activity" or "other"

• "potentially_unintended_data_transfer"

• "potentially_unintended_data_access"

• "potentially_unintended_destructive_activity"

• "other"

• model: string

• object: "safety.alert" (default: "safety.alert")

• "safety.alert"

• reason: string or null

根据 error_type 生成的、适合向客户展示的描述。零数据保留请求中为 null。

• request_id: string

• request_paused: boolean

表示是否成功为该请求登记阻止操作。此字段不确认响应执行是否已经停止。

• response_id: string

• detailed_explanation: optional string or null

生成的解释，仅对符合条件的零数据保留警报临时提供。不可用时省略此字段。

请求示例

## 先设置 OPENAI_API_KEY，并替换文件路径或资源 ID 后再运行。
curl https://api.openai.com/v1/safety/alerts/$ID \
    -H "Authorization: Bearer $OPENAI_API_KEY"

200 响应示例

{
  "id": "id",
  "created_at": 0,
  "error_type": "potentially_unintended_data_transfer",
  "model": "model",
  "object": "safety.alert",
  "reason": "reason",
  "request_id": "request_id",
  "request_paused": true,
  "response_id": "response_id",
  "detailed_explanation": "detailed_explanation"
}

请求示例

## 先设置 OPENAI_API_KEY，并替换文件路径或资源 ID 后再运行。
curl https://api.openai.com/v1/safety/alerts/alert_0123456789abcdef0123456789abcdef \
  -H "Authorization: Bearer $OPENAI_API_KEY"

响应示例

{
  "id": "alert_0123456789abcdef0123456789abcdef",
  "object": "safety.alert",
  "created_at": 1787659200,
  "request_id": "req_123",
  "response_id": "resp_123",
  "model": "gpt-6-astra",
  "request_paused": true,
  "error_type": "potentially_unintended_data_access",
  "reason": "Potentially unintended data access."
}