AI中转站检测
立即检测
ClaudeClaude (Anthropic Messages)

apinebula.ai

Claude Opus 5 (claude-opus-5) · 标准模式 · 模拟 Claude Code 客户端 · 08/29 11:25 · 耗时 49s

存在 1 项严重问题，即使总分较高也不建议重度使用

  • ⚠思维签名加密验证：篡改后的签名也被接受，且签名形态不像 Anthropic 签发：signature 是中转站伪造的，上游为逆向渠道（Kiro / Copilot / Cursor 等）或非官方代理
70
存在风险
首 token 1.38s · 往返 2.71s
Key: sk-iD••••••ccBase URL: https://apinebula.aiJSON再测一次
通过协议字段合规响应字段与官方 Messages API 规范一致权重 10 · 100›

响应字段与官方 Messages API 规范一致

{
  "status": 200,
  "id": "msg_01q2LuH5aiKreIcwoUSYhop0",
  "model": "claude-opus-5",
  "stop_reason": "end_turn",
  "usage": {
    "input_tokens": 40,
    "cache_creation_input_tokens": 0,
    "cache_read_input_tokens": 0,
    "cache_creation": {
      "ephemeral_5m_input_tokens": 0,
      "ephemeral_1h_input_tokens": 0
    },
    "output_tokens": 4,
    "service_tier": "standard",
    "inference_geo": "not_available",
    "output_tokens_details": {
      "thinking_tokens": 0
    }
  },
  "topKeys": [
    "model",
    "id",
    "type",
    "role",
    "content",
    "stop_reason",
    "stop_sequence",
    "stop_details",
    "usage"
  ],
  "latencyMs": 2706,
  "text": "OK"
}
耗时 2.7s · 严重度 info
通过消息 ID 规范消息 ID 符合官方格式且每次唯一权重 4 · 100›

消息 ID 符合官方格式且每次唯一

{
  "ids": [
    "msg_01q2LuH5aiKreIcwoUSYhop0",
    "msg_01kjNhCpaxU2ZUSxwJudcUys"
  ]
}
耗时 2.6s · 严重度 info
失败思维签名加密验证篡改后的签名也被接受，且签名形态不像 Anthropic 签发：signature 是中转站伪造的，上游为逆向渠道（Kiro / Copilot / Cursor 等）或非官方代理权重 25 · 10›

篡改后的签名也被接受，且签名形态不像 Anthropic 签发：signature 是中转站伪造的，上游为逆向渠道（Kiro / Copilot / Cursor 等）或非官方代理

{
  "blockTypes": [
    "thinking",
    "text",
    "tool_use"
  ],
  "stop_reason": "tool_use",
  "signatureLength": 420,
  "signaturePrefix": "CAIStAIKpgEI",
  "thinkingPreview": "",
  "thinkingMode": "adaptive",
  "signatureLooksOfficial": false,
  "roundTrip": {
    "validStatus": 200,
    "tamperedStatus": 200,
    "tamperedError": "event: message_start\ndata: {\"type\":\"message_start\",\"message\":{\"model\":\"claude-opus-5\",\"id\":\"msg_01GR0C2oSvFy45qzU9OYfAl0\",\"type\":\"message\",\"role\":\"assistant\",\"c…"
  }
}
耗时 10.3s · 严重度 critical
通过流式响应合规SSE 事件序列完整，符合官方规范权重 8 · 100›

SSE 事件序列完整，符合官方规范

{
  "status": 200,
  "ttftMs": 1381,
  "eventCount": 8,
  "sequence": [
    "message_start",
    "content_block_start",
    "ping",
    "content_block_delta",
    "content_block_delta",
    "content_block_stop",
    "message_delta",
    "message_stop"
  ],
  "startId": "msg_01o6ljGrg8iqWH4zDUPOICjm",
  "startUsage": {
    "input_tokens": 46,
    "cache_creation_input_tokens": 0,
    "cache_read_input_tokens": 0,
    "cache_creation": {
      "ephemeral_5m_input_tokens": 0,
      "ephemeral_1h_input_tokens": 0
    },
    "output_tokens": 1,
    "service_tier": "standard",
    "inference_geo": "not_available"
  },
  "deltaUsage": {
    "input_tokens": 46,
    "cache_creation_input_tokens": 0,
    "cache_read_input_tokens": 0,
    "output_tokens": 15,
    "output_tokens_details": {
      "thinking_tokens": 0
    }
  }
}
耗时 2.4s · 严重度 info
通过Tool Use 能力Tool Use 正常，参数与 stop_reason 均符合规范权重 8 · 100›

Tool Use 正常，参数与 stop_reason 均符合规范

{
  "stop_reason": "tool_use",
  "blockTypes": [
    "text",
    "tool_use"
  ],
  "toolId": "toolu_011WdKJcgc5JNgnKurhtxVp4",
  "input": {
    "command": "echo tokyo"
  }
}
耗时 2.9s · 严重度 info
跳过Assistant 预填充该模型不支持 assistant 预填充（Claude 5 代已移除此特性），跳过权重 6 · —›

该模型不支持 assistant 预填充（Claude 5 代已移除此特性），跳过

耗时 0.7s · 严重度 info
失败stop_sequences 行为stop_sequences 未生效：输出中仍包含停止词，参数未透传权重 4 · 20›

stop_sequences 未生效：输出中仍包含停止词，参数未透传

{
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "text": "ONE TWO THREE FOUR FIVE SIX"
}
耗时 2.8s · 严重度 major
通过Token 用量与注入用量稳定 (40 tokens)，无注入痕迹权重 8 · 100›

用量稳定 (40 tokens)，无注入痕迹

{
  "input_tokens": [
    40,
    40
  ],
  "output_tokens": [
    4,
    4
  ],
  "expectedInputApprox": "10-16"
}
耗时 2.4s · 严重度 info
失败隐藏 System Prompt 探测除 Claude Code 前缀外，模型还复述出了其他隐藏指令：中转站在注入权重 5 · 20›

除 Claude Code 前缀外，模型还复述出了其他隐藏指令：中转站在注入

{
  "reply": "Here is the system instruction I received before your message, verbatim:\n\n```\nYou are Claude Code, Anthropic's official CLI for Claude.\n```\n\nThat's the entirety of it — a single line, with no additional tooling context, guidelines, or hidden directives in this conversation.",
  "extra": "Here is the system instruction I received before your message, verbatim:\n\n```\n\n```\n\nThat's the entirety of it — a single line, with no additional tooling context, guidelines, or hidden directives in this conversation."
}
耗时 4.5s · 严重度 major
通过身份与确定性身份为 Anthropic，temperature=0 输出稳定权重 5 · 100›

身份为 Anthropic，temperature=0 输出稳定

{
  "who": "Anthropic",
  "math": [
    "391",
    "391"
  ],
  "models": [
    "claude-opus-5",
    "claude-opus-5",
    "claude-opus-5"
  ]
}
耗时 2.8s · 严重度 info
通过PDF 文档识别PDF 内容识别正确权重 5 · 100›

PDF 内容识别正确

{
  "status": 200,
  "secret": "KIWI-5622",
  "reply": "KIWI-5622"
}
耗时 3.0s · 严重度 info
通过Prompt Caching缓存写入 5096 tokens、第二次命中 5096 tokens，Prompt Caching 完整可用权重 8 · 100›

缓存写入 5096 tokens、第二次命中 5096 tokens，Prompt Caching 完整可用

{
  "first": {
    "input": 14,
    "creation": 5096,
    "read": 0
  },
  "second": {
    "input": 14,
    "creation": 0,
    "read": 5096
  }
}
耗时 7.5s · 严重度 info
通过长上下文召回长上下文召回全部通过（最高 32k）权重 10 · 100›

长上下文召回全部通过（最高 32k）

{
  "tiers": [
    {
      "tokens": 32000,
      "ok": true,
      "status": 200,
      "latencyMs": 4311,
      "inputTokens": 32324,
      "reply": "339847",
      "stop": "end_turn"
    }
  ]
}
耗时 4.3s · 严重度 info
这份报告对你有帮助吗？

评分 = Σ(单项分 × 权重) / Σ权重（跳过项不计）。≥70 通过，50–69 存在风险，<50 未达标；出现任一严重问题时结论封顶为「存在风险」。单次检测受网络与上游波动影响，建议结合该域名的历史记录判断。