Talk to your spectrum analyzer.

Claude Code

• Recommended: add this repository as a marketplace and install the plugin (two commands, inside Claude Code or with claude plugin ... in a shell). The plugin runs npx -y uni-t-uts-mcp pinned to this release (0.5.0) and asks for the analyzer address, port and transport (all optional; they are stored in your Claude Code settings).
• Or run the claude mcp add command once in a terminal. Add --scope user to make it available in every project (local is the default, project writes .mcp.json).
• UTS_IP is optional: without it, tell the assistant "connect to the analyzer at 192.168.1.100".

/plugin marketplace add Reid-n0rc/uni-t-uts-mcp
/plugin install uni-t-uts@uni-t-uts-mcp

claude plugin marketplace add Reid-n0rc/uni-t-uts-mcp
claude plugin install uni-t-uts@uni-t-uts-mcp

claude mcp add uni-t-uts -e UTS_IP=192.168.1.100 -- npx -y uni-t-uts-mcp

claude mcp add uni-t-uts -e UTS_IP=192.168.1.100 -- cmd /c npx -y uni-t-uts-mcp

claude mcp add uni-t-uts -- docker run -i --rm -e UTS_IP=192.168.1.100 ghcr.io/reid-n0rc/uni-t-uts-mcp:0.5.0

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "npx",
      "args": [
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

Plugin, marketplace and .mcp.json formats checked against the Claude Code docs and validated with claude plugin validate. claude plugin marketplace add and claude plugin install were run against a local checkout in an isolated Claude Code config directory (install and listing succeeded). Not yet verified: adding the marketplace from GitHub, and starting the server inside a Claude Code session (the pinned npm version must be published first). Source: https://code.claude.com/docs/en/plugins/marketplace-reference.md and https://code.claude.com/docs/en/mcp

Claude Desktop

Download the .mcpb (GitHub releases)

• Easiest: download uni-t-uts-mcp-0.5.0.mcpb from the release and double-click it; the install dialog asks for the analyzer address. (Not verified in Claude Desktop yet.)
• Or edit the config file (Settings > Developer > Edit Config): macOS ~/Library/Application Support/Claude/claude_desktop_config.json, Windows %APPDATA%\Claude\claude_desktop_config.json, Linux ~/.config/Claude/claude_desktop_config.json. Restart Claude Desktop completely afterwards.
• UTS_IP is optional: without it, tell the assistant "connect to the analyzer at 192.168.1.100".

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "npx",
      "args": [
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "cmd",
      "args": [
        "/c",
        "npx",
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "UTS_IP=192.168.1.100",
        "ghcr.io/reid-n0rc/uni-t-uts-mcp:0.5.0"
      ]
    }
  }
}

The claude_desktop_config.json format and paths are from the MCP docs. Installing the .mcpb in Claude Desktop itself is NOT verified yet (issue #86 tracks the manual check). Source: https://modelcontextprotocol.io/docs/develop/connect-local-servers

Cursor

Add to Cursor

• The install link opens Cursor's install dialog with the npx command (no analyzer address; tell the assistant where it is, or add UTS_IP under env afterwards). On Windows use the config file below.
• Or put the JSON in ~/.cursor/mcp.json (all projects) or .cursor/mcp.json (one project).

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "npx",
      "args": [
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "cmd",
      "args": [
        "/c",
        "npx",
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "UTS_IP=192.168.1.100",
        "ghcr.io/reid-n0rc/uni-t-uts-mcp:0.5.0"
      ]
    }
  }
}

File locations and the install-link format are from Cursor's docs (https://cursor.com/docs/mcp/install-links). Not clicked in Cursor by the maintainer. Source: https://cursor.com/docs/context/mcp

VS Code

• Workspace: save the JSON as .vscode/mcp.json. For your user profile run code --add-mcp with the JSON below, or use the Command Palette: MCP: Add Server.
• UTS_IP is optional: without it, tell the assistant "connect to the analyzer at 192.168.1.100".

{
  "servers": {
    "uni-t-uts": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "servers": {
    "uni-t-uts": {
      "type": "stdio",
      "command": "cmd",
      "args": [
        "/c",
        "npx",
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

code --add-mcp '{"name":"uni-t-uts","type":"stdio","command":"npx","args":["-y","uni-t-uts-mcp"],"env":{"UTS_IP":"192.168.1.100"}}'

The servers format, file locations and code --add-mcp are from the VS Code docs (which note the servers format belongs to the Local harness, deprecated from February 2027). No one-click link: the vscode:mcp/install format is not in the official docs, so none is offered. Not run by the maintainer. Source: https://code.visualstudio.com/docs/agents/reference/mcp-configuration

Windsurf

• Open the file with Cascade's Open MCP config file action (or Settings > Tools > Windsurf Settings > Add Server) and add the entry.
• UTS_IP is optional: without it, tell the assistant "connect to the analyzer at 192.168.1.100".

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "npx",
      "args": [
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "cmd",
      "args": [
        "/c",
        "npx",
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "UTS_IP=192.168.1.100",
        "ghcr.io/reid-n0rc/uni-t-uts-mcp:0.5.0"
      ]
    }
  }
}

The mcpServers format is from the Windsurf docs. Their path has moved: the current docs say ~/.config/devin/mcp_config.json (macOS / Linux) or %AppData%/devin/mcp_config.json (Windows), older ones ~/.codeium/windsurf/mcp_config.json. Use Cascade's Open MCP config file action to find the right one. No install link is documented. Not run by the maintainer. Source: https://docs.devin.ai/windsurf/plugins/cascade/mcp

Other stdio clients

• Command npx, arguments -y uni-t-uts-mcp, environment variable UTS_IP (optional, UTS_PORT defaults to 5025).
• On Windows, where npx is a .cmd file, use cmd with arguments /c npx -y uni-t-uts-mcp.

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "npx",
      "args": [
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "cmd",
      "args": [
        "/c",
        "npx",
        "-y",
        "uni-t-uts-mcp"
      ],
      "env": {
        "UTS_IP": "192.168.1.100"
      }
    }
  }
}

{
  "mcpServers": {
    "uni-t-uts": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "UTS_IP=192.168.1.100",
        "ghcr.io/reid-n0rc/uni-t-uts-mcp:0.5.0"
      ]
    }
  }
}

Plain MCP stdio: the client starts the command and talks over stdin/stdout. Key names differ per client; check its docs. Source: https://modelcontextprotocol.io/docs/develop/connect-local-servers