Repository navigation
Add to Client
Copy-paste configuration for each MCP client. Everything here is generated from scripts/client-snippets.mjs (run npm run clients:sync after changing it), so the README, the site and this page agree. 192.168.1.100 is a placeholder: use your analyzer's address. UTS_IP is optional; see Getting Started for the other install forms (release tarball, git tag, a clone).
Source: https://code.claude.com/docs/en/plugins/marketplace-reference.md and https://code.claude.com/docs/en/mcp
Verification: 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).
- 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 runsnpx -y uni-t-uts-mcppinned to this release (1.1.0) and asks for the analyzer address, port and transport (all optional; they are stored in your Claude Code settings). - Or run the
claude mcp addcommand once in a terminal. Add--scope userto make it available in every project (localis the default,projectwrites.mcp.json). - UTS_IP is optional: without it, tell the assistant "connect to the analyzer at 192.168.1.100".
Plugin (recommended):
/plugin marketplace add Reid-n0rc/uni-t-uts-mcp
/plugin install uni-t-uts@uni-t-uts-mcpPlugin from a shell:
claude plugin marketplace add Reid-n0rc/uni-t-uts-mcp
claude plugin install uni-t-uts@uni-t-uts-mcpnpx (macOS / Linux):
claude mcp add uni-t-uts -e UTS_IP=192.168.1.100 -- npx -y uni-t-uts-mcpnpx (Windows):
claude mcp add uni-t-uts -e UTS_IP=192.168.1.100 -- cmd /c npx -y uni-t-uts-mcpDocker:
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:1.1.0.mcp.json (project scope):
{
"mcpServers": {
"uni-t-uts": {
"command": "npx",
"args": [
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}Source: https://modelcontextprotocol.io/docs/develop/connect-local-servers
Verification: 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).
One-click: Download the .mcpb (GitHub releases)
- Easiest: download
uni-t-uts-mcp-1.1.0.mcpbfrom 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".
npx (macOS / Linux):
{
"mcpServers": {
"uni-t-uts": {
"command": "npx",
"args": [
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}npx (Windows):
{
"mcpServers": {
"uni-t-uts": {
"command": "cmd",
"args": [
"/c",
"npx",
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}Docker:
{
"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:1.1.0"
]
}
}
}Source: https://cursor.com/docs/context/mcp
Verification: 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.
One-click: 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_IPunderenvafterwards). On Windows use the config file below. - Or put the JSON in
~/.cursor/mcp.json(all projects) or.cursor/mcp.json(one project).
npx (macOS / Linux):
{
"mcpServers": {
"uni-t-uts": {
"command": "npx",
"args": [
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}npx (Windows):
{
"mcpServers": {
"uni-t-uts": {
"command": "cmd",
"args": [
"/c",
"npx",
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}Docker:
{
"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:1.1.0"
]
}
}
}Source: https://code.visualstudio.com/docs/agents/reference/mcp-configuration
Verification: 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.
- Workspace: save the JSON as
.vscode/mcp.json. For your user profile runcode --add-mcpwith 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".
.vscode/mcp.json (macOS / Linux):
{
"servers": {
"uni-t-uts": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}.vscode/mcp.json (Windows):
{
"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 (macOS / Linux):
code --add-mcp '{"name":"uni-t-uts","type":"stdio","command":"npx","args":["-y","uni-t-uts-mcp"],"env":{"UTS_IP":"192.168.1.100"}}'Source: https://docs.devin.ai/windsurf/plugins/cascade/mcp
Verification: 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.
- 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".
npx (macOS / Linux):
{
"mcpServers": {
"uni-t-uts": {
"command": "npx",
"args": [
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}npx (Windows):
{
"mcpServers": {
"uni-t-uts": {
"command": "cmd",
"args": [
"/c",
"npx",
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}Docker:
{
"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:1.1.0"
]
}
}
}Source: https://modelcontextprotocol.io/docs/develop/connect-local-servers
Verification: Plain MCP stdio: the client starts the command and talks over stdin/stdout. Key names differ per client; check its docs.
- Command
npx, arguments-y uni-t-uts-mcp, environment variableUTS_IP(optional,UTS_PORTdefaults to 5025). - On Windows, where
npxis a.cmdfile, usecmdwith arguments/c npx -y uni-t-uts-mcp.
npx (macOS / Linux):
{
"mcpServers": {
"uni-t-uts": {
"command": "npx",
"args": [
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}npx (Windows):
{
"mcpServers": {
"uni-t-uts": {
"command": "cmd",
"args": [
"/c",
"npx",
"-y",
"uni-t-uts-mcp"
],
"env": {
"UTS_IP": "192.168.1.100"
}
}
}
}Docker:
{
"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:1.1.0"
]
}
}
}MIT licensed. Verified on a UNI-T UTS3036B (firmware V1.04.0040); other models are untested. This wiki is generated from the repository's docs/ folder, which is authoritative: change it there, not here.