Installation

Engram ships as a single daemon. Pick the method that suits your environment — all of them install the same server and MCP endpoint.

Quick install (recommended)

Install the CLI, then run the setup wizard — it clones, builds, and configures everything:

# Install the CLI globally
npm install -g @engram-ai-memory/cli

# Run the setup wizard
engram setup

# Start the server (API + 3D dashboard on :4901)
engram start

That’s it. Open http://localhost:4901 for the dashboard, /docs for Swagger UI. State is stored in ~/.engram/.

Management commands

engram setup       # First-time setup wizard
engram start       # Start server (background)
engram stop        # Stop the server
engram doctor      # Health checks (Node, pnpm, DB, MCP, API)
engram status      # Server status + memory count
engram configure   # View/set config (port, dbPath, namespace)

Shell script install (alternative)

# macOS / Linux
curl -fsSL https://engram.am/install.sh | bash

# Windows (PowerShell)
powershell -c "irm https://engram.am/install.ps1 | iex"

Build from source (development)

git clone https://github.com/ayvazyan10/engram.git
cd engram
pnpm install
pnpm turbo run build
node apps/server/dist/index.js

Start the API server

The Engram API server provides REST endpoints, WebSocket events, and Swagger UI:

cd engram
ENGRAM_DB_PATH=./engram.db node apps/server/dist/index.js

The server starts on port 4901 by default. Key environment variables:

See Configuration for all environment variables.

Open the 3D dashboard

After building, the 3D dashboard is served automatically from the API server — no separate process needed:

# Dashboard is at http://localhost:4901 (same as API)
# Just start the server after building:
node apps/server/dist/index.js

The dashboard shows all memories as an interactive 3D neural graph with 5 visual modes — Cosmos, Nebula, Neural Net, Galaxy, and Clusters. Swagger UI is at http://localhost:4901/docs.

Verify the server

curl http://localhost:4901/api/health
# → {"status":"ok","version":"0.3.1","uptime":12}

Store your first memory

# Store
curl -s -X POST http://localhost:4901/api/memory \
  -H "Content-Type: application/json" \
  -d '{"content":"User prefers TypeScript","type":"semantic","importance":0.8}'

# Recall
curl -s -X POST http://localhost:4901/api/recall \
  -H "Content-Type: application/json" \
  -d '{"query":"What language does the user prefer?","maxTokens":2000}'

Or use the CLI:

engram store "User prefers TypeScript" --type semantic --importance 0.8
engram recall "What language does the user prefer?" --raw

Connect to Claude Code

Add Engram as an MCP server in ~/.mcp.json:

{
  "mcpServers": {
    "engram": {
      "command": "node",
      "args": ["/path/to/engram/packages/mcp/dist/server.js"],
      "env": {
        "ENGRAM_DB_PATH": "/path/to/engram.db"
      }
    }
  }
}

Restart Claude Code. Engram’s 21 tools will appear automatically:

See MCP Integration for the full setup guide and MCP Tools for parameter reference.

Swagger UI

Interactive API documentation is available at http://localhost:4901/docs once the server is running. All 40+ endpoints are documented with request/response schemas.