VS Code

VS Code does not support MCP natively, but several extensions add MCP client capabilities. Once configured, your AI assistant has persistent memory across sessions.

Prerequisites

VS Code installed
An MCP-compatible extension: Continue, Cline, Roo Code, or similar

Installation

The installer can detect and configure most MCP extensions automatically:

curl -fsSL https://get.engrammic.ai/install.sh | sh
engrammic install

If your extension is not detected, configure it manually below.

Manual Configuration

Config location varies by extension. Find your extension below and add the Engrammic server.

Create .continue/mcpServers/engrammic.yaml in your project:

mcpServers:
  - name: engrammic
    type: streamable-http
    url: http://localhost:8000/mcp

Create .cline/mcp.json in your project:

{
  "mcpServers": {
    "engrammic": {
      "url": "http://localhost:8000/mcp"
    }
  }
}

Check your extension's documentation for MCP configuration. The server config is:

{
  "engrammic": {
    "type": "http",
    "url": "http://localhost:8000/mcp"
  }
}

Using Engrammic Cloud? Replace http://localhost:8000/mcp with https://beta.engrammic.ai/mcp/ in the config above. Run engrammic-mcp login to authenticate.

Verify

Restart VS Code after configuration. Open your extension's chat and ask:

What memory tools do you have available?

You should see tools like remember, learn, recall, and trace.

Load the EAG Guide

Engrammic includes a workflow guide that teaches your agent when and how to use memory tools effectively. After verifying the connection, have your agent load it:

Load the onboarding pattern from Engrammic

For the full guide, see EAG Guide.

Next Steps

EAG Guide - understand the memory model
Troubleshooting - common issues and fixes
Architecture - how the layers work

Other pages in the docs that reference this one:

IntegrationsAdd Engrammic memory to your AI coding tools