agent-memory
Cloudflare Workers MCP server: agent-memory
我该使用它吗
质量与安全性
A
发现(1)
- LOW在 store_memory 中
基于对工具定义和协议合规性的自动分析。
上下文开销
~331token 数(工具定义)
~588 B典型响应大小
对注意力的影响极小(占 128k 上下文窗口的 0.26%)
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": {
"agent-memory": {
"url": "https://api.lazy-mac.com/agent-memory/mcp"
}
}
}远程端点
https://api.lazy-mac.com/agent-memory/mcpstreamable-http它能做什么
工具清单
工具(5)
🟢 只读🟡 写入🔴 删除⚪ 未知
⚪store_memory(namespace, content, metadata, ttl)
Store a memory for an AI agent. Persists across sessions.
输入模式
{
"type": "object",
"properties": {
"namespace": {
"type": "string",
"description": "Namespace to group memories (e.g. agent name)"
},
"content": {
"type": "string",
"description": "Memory content to store"
},
"metadata": {
"type": "object",
"description": "Optional metadata key-value pairs"
},
"ttl": {
"type": "number",
"description": "Optional TTL in seconds"
}
},
"required": [
"namespace",
"content"
]
}🟢search_memories(namespace, q, limit)
Search memories by namespace with optional text query.
输入模式
{
"type": "object",
"properties": {
"namespace": {
"type": "string",
"description": "Namespace to search in"
},
"q": {
"type": "string",
"description": "Optional search query (substring match)"
},
"limit": {
"type": "number",
"description": "Max results to return (default 50, max 200)"
}
},
"required": [
"namespace"
]
}🟢get_memory(id)
Retrieve a specific memory by ID.
输入模式
{
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "Memory ID"
}
},
"required": [
"id"
]
}🔴delete_memory(id)
Delete a specific memory by ID.
输入模式
{
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "Memory ID"
}
},
"required": [
"id"
]
}🔴clear_namespace(namespace)
Delete all memories in a namespace.
输入模式
{
"type": "object",
"properties": {
"namespace": {
"type": "string",
"description": "Namespace to clear"
}
},
"required": [
"namespace"
]
}社区
证据
最近观测
已验证未记录版本5 个工具
已验证未记录版本5 个工具