MCP MeasureMCP Measure

索引已觀測

past.dev

Memory for AI agents. Ask what is true now, get the facts back with their dated sources.

我該用這個嗎

品質與安全性

A
說明品質
100%
結構描述完整度
77%
命名品質
100%
汙染風險
100%
權限相符程度
100%
協定合規性
100%

根據工具定義與協定合規性的自動化分析。

上下文成本

~286Token（工具定義）
~459 B典型回應大小
極小的注意力影響（128k 上下文的 0.22%）

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": {
    "past": {
      "url": "https://past.dev/mcp"
    }
  }
}

遠端端點

https://past.dev/mcpstreamable-http
https://app.past.dev/mcpstreamable-http

它能做什麼

工具清單

工具（3）

🟢 唯讀🟡 寫入🔴 刪除⚪ 未知
🟢search_docs(query, limit)

Search the past.dev developer documentation (Memory API and MCP server) and get back matching pages with an excerpt. Lexical search over the full corpus.

輸入結構描述

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "description": "What to look for, in plain language."
    },
    "limit": {
      "type": "integer",
      "description": "Max results, default 5."
    }
  },
  "required": [
    "query"
  ]
}
🟢read_page(product, slug)

Read one documentation page as Markdown. Get product and slug values from list_pages or search_docs.

輸入結構描述

{
  "type": "object",
  "properties": {
    "product": {
      "type": "string",
      "description": "\"memory-api\" or \"mcp\"."
    },
    "slug": {
      "type": "string",
      "description": "Page slug, e.g. \"quickstart\" or \"api-reference\"."
    }
  },
  "required": [
    "product",
    "slug"
  ]
}
🟢list_pages

List every documentation page with its product, slug, title and URL.

輸入結構描述

{
  "type": "object",
  "properties": {}
}

社群

為此伺服器評分

證據

近期觀測

已驗證未記錄版本3 個工具