Trust Live
Evidence-linked MCP selection and non-executing Safe-Call review; surfaces unknowns.
我该使用它吗
质量与安全性
A
发现（1）
- LOWTool 'compare_providers' description lacks action verb在 compare_providers 中
基于对工具定义和协议合规性的自动分析。
上下文开销
~735token 数（工具定义）
~1.2 KB典型响应大小
对注意力有中等影响（占 128k 上下文窗口的 0.57%）
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": {
"trust-live": {
"url": "https://mcp-selection-lab-production.up.railway.app/trust-live/mcp"
}
}
}远程端点
https://mcp-selection-lab-production.up.railway.app/trust-live/mcpstreamable-http它能做什么
工具清单
工具（5）
🟢 只读🟡 写入🔴 删除⚪ 未知
🟢recommend_provider(task_requirements)
Recommend a provider with hard-filter outcomes, component scores, evidence, and Safe-Call constraints.
输入模式
{
"type": "object",
"properties": {
"task_requirements": {
"additionalProperties": true,
"description": "Structured capability, operation, auth, payment, data and evidence requirements.",
"title": "Task Requirements",
"type": "object"
}
},
"required": [
"task_requirements"
],
"title": "recommend_providerArguments"
}输出模式
{
"type": "object",
"additionalProperties": true,
"title": "recommend_providerDictOutput"
}🟢compare_providers(task_requirements)
Compare matching providers across capability, verification, safety, payment, freshness and unknowns.
输入模式
{
"type": "object",
"properties": {
"task_requirements": {
"additionalProperties": true,
"description": "Structured task requirements.",
"title": "Task Requirements",
"type": "object"
}
},
"required": [
"task_requirements"
],
"title": "compare_providersArguments"
}输出模式
{
"type": "object",
"additionalProperties": true,
"title": "compare_providersDictOutput"
}🟢get_provider_profile(provider_id)
Get a canonical provider profile with capabilities, evidence references, and safe-call contracts.
输入模式
{
"type": "object",
"properties": {
"provider_id": {
"maxLength": 128,
"minLength": 1,
"title": "Provider Id",
"type": "string"
}
},
"required": [
"provider_id"
],
"title": "get_provider_profileArguments"
}输出模式
{
"type": "object",
"additionalProperties": true,
"title": "get_provider_profileDictOutput"
}🟢get_capability(capability_id)
Get every provider capability with this canonical action ID, including evidence and Safe-Call contract.
输入模式
{
"type": "object",
"properties": {
"capability_id": {
"maxLength": 128,
"minLength": 1,
"title": "Capability Id",
"type": "string"
}
},
"required": [
"capability_id"
],
"title": "get_capabilityArguments"
}输出模式
{
"type": "object",
"additionalProperties": true,
"title": "get_capabilityDictOutput"
}🟢evaluate_safe_call(provider_id, capability_id, arguments, user_confirmed, repository_visibility, ...)
Evaluate Safe-Call constraints without invoking the provider tool.
输入模式
{
"type": "object",
"properties": {
"provider_id": {
"maxLength": 128,
"minLength": 1,
"title": "Provider Id",
"type": "string"
},
"capability_id": {
"maxLength": 128,
"minLength": 1,
"title": "Capability Id",
"type": "string"
},
"arguments": {
"additionalProperties": true,
"description": "Proposed arguments; evaluated locally and never sent to the provider.",
"title": "Arguments",
"type": "object"
},
"user_confirmed": {
"default": false,
"title": "User Confirmed",
"type": "boolean"
},
"repository_visibility": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Repository Visibility"
},
"data_classes": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Data Classes"
}
},
"required": [
"provider_id",
"capability_id",
"arguments"
],
"title": "evaluate_safe_callArguments"
}输出模式
{
"type": "object",
"additionalProperties": true,
"title": "evaluate_safe_callDictOutput"
}社区
证据
最近观测
已验证未记录版本5 个工具