Skip to main content

​
MCP Integration

Remem provides a Model Context Protocol (MCP) server that lets AI agents query your context directly.

​
Available Tools

ToolDescriptionParameters
remem_queryQuery for relevant contextquery (string), mode (fast/rich), max_results (int)
remem_ingestIngest a new documenttitle (string), content (string), metadata (object)
remem_searchSimple keyword searchq (string), limit (int)

​
Claude Desktop Configuration

Add Remem to your Claude Desktop claude_desktop_config.json: 
{
  "mcpServers": {
    "remem": {
      "command": "uvx",
      "args": ["remem-mcp"],
      "env": {
        "REMEM_API_URL": "https://api.remem.io",
        "REMEM_API_KEY": "vlt_your_key_here"
      }
    }
  }
}

​
Usage in Claude

Once configured, Claude can automatically use Remem tools:
You: What did we discuss in last week’s planning meeting? Claude: Uses remem_query to search for meeting notes… Let me check your notes. Based on your planning meeting from January 15th, you discussed three priorities:
  1. Expand to EU markets
  2. Launch the mobile app by March
  3. Hire two more engineers

​
Environment Variables

VariableRequiredDescription
REMEM_API_URLYesBase URL of your Remem instance
REMEM_API_KEYYesAPI key with appropriate sensitivity scope
REMEM_DEFAULT_MODENoDefault query mode (fast or rich, defaults to fast)
REMEM_MAX_RESULTSNoDefault max results (defaults to 10)

​
Security

The MCP server runs locally and communicates with the Remem API using your API key. The key’s sensitivity scope controls which documents the agent can access — use a public or internal scoped key if you want to limit what the agent can see.