agentram-mcp
Provides persistent memory for AI agents via 10 MCP tools that map to the AgentRAM REST API, enabling store, retrieve, search, and share memories across personal and shared namespaces.
README
agentram-mcp
Model Context Protocol server for AgentRAM. Give your AI agents persistent memory in Claude Desktop, Cline, Cursor, and any other MCP-compatible client.
What you get
10 MCP tools that map 1:1 to the AgentRAM REST API.
Personal agent memory
store_memory: Write a value under anagent_id+key. Optionalttl_daysfor auto-expiry.retrieve_memory: Read a memory back byagent_id+key. Refunds the credit on 404.list_memories: List everything stored for an agent, newest first.search_memories: Case-insensitive text search across keys and values. No embeddings needed.delete_memory: Permanently delete a memory.
Shared multi-agent memory
create_namespace: Create a shared memory pool that multiple agents can read and write to. Free.store_shared_memory: Write to a shared namespace.retrieve_shared_memory: Read from a shared namespace.list_shared_memories: List everything in a namespace.
Utility
check_credits: Show your current account balance. Free, no credit deducted.
Each non-free tool costs 1 credit. Reads and writes are the same price.
Get an API key
Sign up at agentram.dev. You get 100 free credits on signup. No credit card required.
Install in Claude Desktop
Edit claude_desktop_config.json:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
Add:
{
"mcpServers": {
"agentram": {
"command": "npx",
"args": ["-y", "agentram-mcp"],
"env": {
"AGENTRAM_API_KEY": "agentram_your_key_here"
}
}
}
}
Restart Claude Desktop. The 10 tools will appear in the MCP picker.
Install in Cline (VS Code)
In Cline's MCP settings, add:
{
"mcpServers": {
"agentram": {
"command": "npx",
"args": ["-y", "agentram-mcp"],
"env": {
"AGENTRAM_API_KEY": "agentram_your_key_here"
}
}
}
}
Install in Cursor
In Cursor's MCP settings, add:
{
"mcpServers": {
"agentram": {
"command": "npx",
"args": ["-y", "agentram-mcp"],
"env": {
"AGENTRAM_API_KEY": "agentram_your_key_here"
}
}
}
}
Try it locally before publishing
npm install
npm run build
AGENTRAM_API_KEY=your_key node dist/index.js
Or open the MCP Inspector web UI:
AGENTRAM_API_KEY=your_key npx @modelcontextprotocol/inspector node dist/index.js
The Inspector lets you list tools and invoke each one manually to verify everything works against the real API.
Environment variables
| Variable | Required | Description |
|---|---|---|
AGENTRAM_API_KEY |
Yes | Your AgentRAM API key, starts with agentram_ |
AGENTRAM_API_BASE |
No | Override the API base URL. Defaults to https://api.agentram.dev |
Links
- AgentRAM website: agentram.dev
- Full API docs: agentram.dev/docs.html
- Issues: github.com/seanmarkwei/agentram-mcp/issues
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。