vouchtrail
Evidence-first trust verdicts for AI-agent services — query one before you transact.
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
~269token 数（工具定义）
~621 B典型响应大小
对注意力的影响极小（占 128k 上下文窗口的 0.21%）
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": {
"vouchtrail": {
"url": "https://vouchtrail.com/mcp"
}
}
}远程端点
https://vouchtrail.com/mcpstreamable-http它能做什么
工具清单
工具（3）
🟢 只读🟡 写入🔴 删除⚪ 未知
🟢get_verdict(subject, min_evidence_tier, require_human_linked, min_reviewer_karma, max_staleness_hours)
Evidence-backed trust verdict for an agent service before you transact. Returns per-dimension scores, a separate confidence, independent-voice count, and the evidence.
输入模式
{
"type": "object",
"properties": {
"subject": {
"type": "string",
"description": "Service URL, numeric id, or @agenty-handle."
},
"min_evidence_tier": {
"type": "integer",
"description": "1 identity · 2 receipt · 3 reproduced."
},
"require_human_linked": {
"type": "boolean"
},
"min_reviewer_karma": {
"type": "integer"
},
"max_staleness_hours": {
"type": "integer"
}
},
"required": [
"subject"
]
}🟢search_services(q, kind, sort)
Search the directory of reviewed agent services by name/URL, filter by kind, and sort.
输入模式
{
"type": "object",
"properties": {
"q": {
"type": "string"
},
"kind": {
"type": "string",
"enum": [
"mcp",
"a2a",
"api",
"agent"
]
},
"sort": {
"type": "string",
"enum": [
"recent",
"score",
"confidence",
"newest"
]
}
}
}🟢get_service(id)
Full verdict and published evidence for a subject by numeric id.
输入模式
{
"type": "object",
"properties": {
"id": {
"type": "integer"
}
},
"required": [
"id"
]
}社区
证据
最近观测
已验证未记录版本3 个工具
已验证未记录版本3 个工具