Connect any MCP client to HyperMemory
HyperMemory is a standards-compliant MCP server. If your agent or framework supports the Model Context Protocol, it works with HyperMemory.
Why This Works
The Model Context Protocol is becoming the standard for connecting AI agents to external tools and data. HyperMemory implements the MCP server specification with HTTP/SSE transport and OAuth 2.1 authentication. Any client that speaks MCP can connect — no custom adapters, no SDK dependencies, no wrapper code.
Quick Setup
Point your MCP client to the HyperMemory server:
Server URL: https://api.hypermemory.io/mcp
Transport: HTTP/SSE (Streamable HTTP)
Auth: Bearer token (API key)Standard MCP client configuration:
{
"mcpServers": {
"hypermemory": {
"type": "url",
"url": "https://api.hypermemory.io/mcp",
"headers": {
"Authorization": "Bearer YOUR_API_KEY"
}
}
}
}Available Tools
| Tool | Description |
|---|---|
| memory_store | Store a memory node with metadata and relationships |
| memory_recall | Query memories by natural language |
| memory_find_related | Find nodes related to a given node |
| memory_get_relationships | Get all relationships for a node |
| memory_update | Update an existing memory node |
| memory_forget | Remove a memory node |
| memory_export_subgraph | Export a portion of the graph |
| memory_load_link | Import a subgraph from another source |
Use Case
A team building a custom agent framework in Rust adds long-term memory in an afternoon. They point their MCP client at HyperMemory's endpoint, authenticate with an API key, and their agent immediately has access to structured memory tools. No SDK to install, no library to maintain — just the MCP standard.