IDE Setup

IDE Setup

SuperLocalMemory provides documented MCP configurations for named clients. Run slm connect --list to see the current release surface; other MCP-compatible clients can be configured manually but are not automatically release-verified.

Auto-Detection (Recommended)

slm connect        # Auto-detect and configure all installed IDEs
slm connect --list # See which IDEs are configured

After running, restart your IDE to activate the connection.

slm connect writes only product-authored connection and memory protocol to IDE instruction files. It does not copy recalled topics, decisions, entities, or memory text into those high-trust files. Dynamic context is fetched through MCP at runtime and is delivered as bounded, untrusted evidence.

Manual Setup by IDE

If auto-detection does not find your IDE, configure it manually. All IDEs use the same MCP server command:

{
  "mcpServers": {
    "superlocalmemory": {
      "command": "slm",
      "args": ["mcp"]
    }
  }
}

Below are the config file paths for each IDE.

Claude Code

Config file: ~/.claude.json (or project-level .mcp.json)

{
  "mcpServers": {
    "superlocalmemory": {
      "command": "slm",
      "args": ["mcp"]
    }
  }
}

Cursor

Config file: ~/.cursor/mcp.json

{
  "mcpServers": {
    "superlocalmemory": {
      "command": "slm",
      "args": ["mcp"]
    }
  }
}

VS Code (with Copilot MCP extension)

Config file: ~/.vscode/mcp.json

{
  "mcpServers": {
    "superlocalmemory": {
      "command": "slm",
      "args": ["mcp"]
    }
  }
}

Windsurf

Config file: ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "superlocalmemory": {
      "command": "slm",
      "args": ["mcp"]
    }
  }
}

Gemini CLI

Config file: ~/.gemini/settings.json

{
  "mcpServers": {
    "superlocalmemory": {
      "command": "slm",
      "args": ["mcp"]
    }
  }
}

JetBrains IDEs (IntelliJ, PyCharm, WebStorm, etc.)

Settings > Tools > AI Assistant > MCP Servers:

• Name: superlocalmemory
• Command: slm
• Args: mcp

Continue.dev

Config file: ~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "superlocalmemory",
      "command": "slm",
      "args": ["mcp"]
    }
  ]
}

Zed

Config file: ~/.config/zed/settings.json

{
  "language_models": {
    "mcp_servers": {
      "superlocalmemory": {
        "command": "slm",
        "args": ["mcp"]
      }
    }
  }
}

Verifying the Connection

After configuring your IDE:

slm status        # Check SLM is running
slm connect --list # See which IDEs are configured

In your IDE, try asking your AI: "What do you know about my recent work?" If SuperLocalMemory is connected and has stored memories, the AI will reference them.

Troubleshooting

IDE does not detect SuperLocalMemory:

• Verify installation: which slm should return a path
• Verify the MCP server starts: slm mcp (should hang waiting for stdio input — Ctrl+C to stop)
• Check your IDE's MCP config file path is correct

Memories not appearing in AI responses:

• Check that you have stored memories: slm recall "test"
• Check the active profile: slm profile list
• Restart your IDE after config changes

Multiple IDEs: All IDEs share the same memory database. A memory stored via Claude Code is available in Cursor, VS Code, and every other connected IDE.

---

Part of Qualixar | Created by Varun Pratap Bhardwaj