oathlayer
Verify AI outputs against real sources. Returns a trust signal and permanent doc number.
我该使用它吗
质量与安全性
A
基于对工具定义和协议合规性的自动分析。
上下文开销
~118token 数(工具定义)
~810 B典型响应大小
对注意力的影响极小(占 128k 上下文窗口的 0.09%)
这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高,可用于其他任务的注意力就越少。
安装
一键安装
将以下内容添加到你的 `claude_desktop_config.json` 文件中:
{
"mcpServers": {
"oathlayer": {
"url": "https://mcp.oathlayer.com"
}
}
}远程端点
https://mcp.oathlayer.comstreamable-http它能做什么
工具清单
工具(1)
🟢 只读🟡 写入🔴 删除⚪ 未知
🟡verify_output(question, apiKey)
Submit a question to Oathlayer for verification. Returns whether the answer is approved, unverified, reflects multiple perspectives, is too complex to verify, or needs clarification. Also returns a permanent doc reference number.
输入模式
{
"type": "object",
"properties": {
"question": {
"type": "string",
"description": "The question or statement you want Oathlayer to verify."
},
"apiKey": {
"type": "string",
"description": "Your Oathlayer API key from your dashboard at oathlayer.com."
}
},
"required": [
"question",
"apiKey"
]
}社区
证据
最近观测
已验证未记录版本1 个工具
已验证未记录版本1 个工具