Yumi Cost-Aware Diagnostic Planner
Exact expected-cost diagnostic policies for caller-supplied finite deterministic fault models.
我該用這個嗎
品質與安全性
A
根據工具定義與協定合規性的自動化分析。
上下文成本
~291Token（工具定義）
~2.5 KB典型回應大小
極小的注意力影響（128k 上下文的 0.23%）
這是每次將伺服器的工具載入模型上下文時所消耗的約略 token 數量。數量越高，可用於其他工作的注意力就越少。
安裝
一鍵安裝
將以下內容加入你的 `claude_desktop_config.json` 檔案：
{
"mcpServers": {
"yumi-cost-aware-diagnostic-planner": {
"url": "https://noqt.no-qt.chatgpt.site/api/yumi/probe-mcp"
}
}
}遠端端點
https://noqt.no-qt.chatgpt.site/api/yumi/probe-mcpstreamable-http它能做什麼
工具清單
工具（1）
🟢 唯讀🟡 寫入🔴 刪除⚪ 未知
🟢plan_diagnostic_checks(hypotheses, probes)
Plans a finite adaptive diagnostic policy and expected costs from caller-supplied hypotheses, relative weights, probes, costs, and complete deterministic outcomes; it does not infer real-world truth, timing, or costs.
輸入結構描述
{
"type": "object",
"properties": {
"hypotheses": {
"type": "array",
"minItems": 1,
"maxItems": 12,
"items": {
"type": "object",
"properties": {
"id": {
"type": "string",
"minLength": 1,
"maxLength": 64
},
"weight": {
"type": "integer",
"minimum": 1,
"maximum": 1000000
}
},
"required": [
"id",
"weight"
],
"additionalProperties": false
}
},
"probes": {
"type": "array",
"minItems": 0,
"maxItems": 12,
"items": {
"type": "object",
"properties": {
"id": {
"type": "string",
"minLength": 1,
"maxLength": 64
},
"cost": {
"type": "integer",
"minimum": 1,
"maximum": 1000000
},
"outcomes": {
"type": "object",
"minProperties": 1,
"description": "Complete deterministic mapping from every hypothesis id to one outcome string",
"additionalProperties": {
"type": "string",
"minLength": 1,
"maxLength": 64
}
}
},
"required": [
"id",
"cost",
"outcomes"
],
"additionalProperties": false
}
}
},
"required": [
"hypotheses",
"probes"
],
"additionalProperties": false
}社群
證據
近期觀測
已驗證未記錄版本1 個工具