HumanMirror Nexus
Outcome-first agent fallback: free discovery, minimal routing, declared costs, verified execution.
我该使用它吗
质量与安全性
基于对工具定义和协议合规性的自动分析。
上下文开销
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": {
"nexus": {
"url": "https://humanmirror.fr/api/nexus/mcp/"
}
}
}远程端点
https://humanmirror.fr/api/nexus/mcp/streamable-http它能做什么
工具清单
工具（6）
🟢outcome_quote(objective, outcome_type, input, criteria, options)
Free quote for a verifiable HumanMirror Outcome contract. Describe the objective or provide an outcome_type; the quote returns the deterministic verifier, selected internal tool and success-only credit cost.
输入模式
{
"type": "object",
"properties": {
"objective": {
"type": "string",
"description": "Natural-language objective, for example: clean this JSON and make the result stable."
},
"outcome_type": {
"type": "string",
"description": "Optional explicit outcome type returned by a previous quote."
},
"input": {
"description": "JSON input for the requested outcome."
},
"criteria": {
"type": "object",
"description": "Optional measurable success threshold."
},
"options": {
"type": "object",
"description": "Optional operation-specific options."
}
},
"required": [
"input"
],
"additionalProperties": false
}⚪outcome_run(objective, outcome_type, contract_id, input, criteria, ...)
Execute a HumanMirror Outcome contract. Credits are charged only when deterministic verification succeeds; failed execution or failed verification is automatically refunded.
输入模式
{
"type": "object",
"properties": {
"objective": {
"type": "string"
},
"outcome_type": {
"type": "string"
},
"contract_id": {
"type": "string",
"description": "Optional contract_id returned by outcome_quote. If supplied it must match the current input, criteria and options."
},
"input": {
"description": "JSON input for the requested outcome."
},
"criteria": {
"type": "object"
},
"options": {
"type": "object"
}
},
"required": [
"input"
],
"additionalProperties": false
}⚪outcome_verify(outcome_type, input, output, criteria, options)
Free deterministic re-verification of an Outcome result without consuming Nexus credits.
输入模式
{
"type": "object",
"properties": {
"outcome_type": {
"type": "string"
},
"input": {
"description": "Original input."
},
"output": {
"description": "Outcome output to verify."
},
"criteria": {
"type": "object"
},
"options": {
"type": "object"
}
},
"required": [
"outcome_type",
"input",
"output"
],
"additionalProperties": false
}🟢nexus_search(query)
Search the HumanMirror Nexus catalog for a machine-callable tool that matches a task. This search is free and does not consume credits.
输入模式
{
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "Describe the operation the agent needs."
}
},
"required": [
"query"
],
"additionalProperties": false
}⚪magnet_resolve(query)
Free universal fallback discovery. Call this when the agent cannot satisfy a task with its current capabilities. Magnet returns a matching HumanMirror tool or records an aggregated capability gap in the Machine Demand Graph. This tool consumes no credits.
输入模式
{
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "Describe the missing capability or task the agent needs to complete."
}
},
"required": [
"query"
],
"additionalProperties": false
}🟢nexus_call(tool_id, input, options)
Route a tool call through HumanMirror Nexus. Use nexus_search first, then call the selected tool_id. Successful calls consume the number of Nexus credits declared by the selected tool.
输入模式
{
"type": "object",
"properties": {
"tool_id": {
"type": "string",
"description": "A tool_id returned by nexus_search, for example forge.clean_json."
},
"input": {
"description": "JSON input forwarded to the selected tool."
},
"options": {
"type": "object",
"description": "Optional tool-specific options."
}
},
"required": [
"tool_id",
"input"
],
"additionalProperties": false
}社区
证据