DeepMem

Mem0-compatible persistent memory for AI agents: write facts once, recall them semantically.

我该使用它吗

质量与安全性

A
描述质量
100%
模式完整度
77%
命名质量
80%
投毒风险
100%
权限匹配度
100%
协议合规性
100%

基于对工具定义和协议合规性的自动分析。

上下文开销

~410token 数(工具定义)
~1.4 KB典型响应大小
对注意力的影响极小(占 128k 上下文窗口的 0.32%)

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

远程端点

https://deepmem.dev/mcpstreamable-http

它能做什么

工具清单

工具(2)

🟢 只读🟡 写入🔴 删除⚪ 未知
🟡deepmem_write(messages, user_id, infer, agent_id, run_id, ...)

Write conversation messages to DeepMemory for fact extraction and persistent storage. Messages are processed by an LLM to extract structured memories, which are then embedded and stored in a vector database for later semantic search. Set infer=True to enable LLM fact extraction (produces richer memories but costs one LLM call). Set infer=False to store raw messages without extraction. Returns a list of memory IDs for successfully stored facts.

输入模式

{
  "type": "object",
  "properties": {
    "messages": {
      "items": {
        "additionalProperties": true,
        "type": "object"
      },
      "title": "Messages",
      "type": "array"
    },
    "user_id": {
      "default": "default",
      "title": "User Id",
      "type": "string"
    },
    "infer": {
      "default": true,
      "title": "Infer",
      "type": "boolean"
    },
    "agent_id": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Agent Id"
    },
    "run_id": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Run Id"
    },
    "api_key": {
      "default": "",
      "title": "Api Key",
      "type": "string"
    }
  },
  "required": [
    "messages"
  ],
  "title": "deepmem_writeArguments"
}
🟢deepmem_search(query, user_id, top_k, threshold, api_key)

Search memories stored in DeepMemory using semantic search. Returns the most relevant memories for the given query, ranked by hybrid scoring (vector similarity + BM25 keyword match + entity boost + time decay). Use this to retrieve context from past conversations before responding to the user. Memories are scoped to the user_id provided during write.

输入模式

{
  "type": "object",
  "properties": {
    "query": {
      "title": "Query",
      "type": "string"
    },
    "user_id": {
      "default": "default",
      "title": "User Id",
      "type": "string"
    },
    "top_k": {
      "default": 10,
      "title": "Top K",
      "type": "integer"
    },
    "threshold": {
      "default": 0.3,
      "title": "Threshold",
      "type": "number"
    },
    "api_key": {
      "default": "",
      "title": "Api Key",
      "type": "string"
    }
  },
  "required": [
    "query"
  ],
  "title": "deepmem_searchArguments"
}

社区

评价此服务器

证据

最近观测

已验证未记录版本2 个工具
已验证未记录版本2 个工具
已验证未记录版本2 个工具