KY
Developer Tools

KYC MCP Server: Cross-Platform AI Diagnostics

Diagnose systems with the MCP server: cross-platform AI diagnostics and recommendations for Windows, macOS and Linux, compatible with Claude Desktop.

View on GitHub
Quick Install
npx -y @vishnurudra-ai/KYC-mcp-server

Overview

KYC MCP Server is a lightweight Model Context Protocol (MCP) server designed to surface cross-platform system diagnostics for AI agents and assistant integrations (for example, Claude Desktop). It runs locally on Windows, macOS and Linux to collect telemetry and actionable recommendations—system information, process and service status, disk and network health—then exposes those as tools that an AI can call to diagnose issues and propose fixes.

The server is useful when you want an assistant to reason about the host system in a structured, auditable way. Instead of letting an LLM guess at system state, the MCP server gives the model reliable, machine-collected data and predefined actions. This improves troubleshooting accuracy, reduces back-and-forth, and enables reproducible remediation steps while keeping the execution surface explicit and reviewable.

Features

  • Cross-platform support: runs on Windows, macOS and Linux
  • MCP-compatible interface for AI assistants (Claude Desktop compatible)
  • Collection of diagnostics: system info, processes, services, disk usage, network checks, logs
  • Actionable recommendations and guided remediation steps
  • REST and local IPC endpoints for tool invocation
  • Configurable security: origin/host allowlists and API keys
  • Easy to extend with custom tools and scripts

Installation / Configuration

Clone the repository and install dependencies. The project uses Node.js (v16+) in the examples below; adapt if your environment differs.

  1. Clone and install
git clone https://github.com/vishnurudra-ai/KYC-mcp-server.git
cd KYC-mcp-server
npm install
  1. Copy and edit environment configuration
cp .env.example .env
# Edit .env to set MCP_PORT, ALLOWED_ORIGINS, API_KEY, MODE, LOG_LEVEL, etc.

Example .env entries

MCP_PORT=3210
ALLOWED_ORIGINS=http://localhost:3000,http://127.0.0.1:3000
API_KEY=changeme
LOG_LEVEL=info
  1. Run the server
npm run start
# Or for development
npm run dev

Register the MCP server in Claude Desktop (or your MCP-capable client) using the local URL and API key:

  • URL: http://localhost:3210/mcp
  • API key: value from .env

Platform service examples (Linux systemd)

# /etc/systemd/system/kyc-mcp.service
[Unit]
Description=KYC MCP Server
After=network.target

[Service]
Type=simple
User=svcuser
WorkingDirectory=/opt/KYC-mcp-server
ExecStart=/usr/bin/node server.js
Restart=on-failure

[Install]
WantedBy=multi-user.target

Available Tools / Resources

The server exposes a set of built-in tools that the AI can call via MCP. Typical tools include:

Tool namePurpose
sysinfoBasic OS, CPU, memory and uptime
processesList running processes and resource usage
servicesCheck status of system services (systemd/launchd/Windows services)
disk-usageDisk mounts and free space, large file detection
network-checkPing, traceroute, DNS resolution checks
port-scanLocal port occupancy scan for troubleshooting conflicts
collect-logsGather relevant logs (by path or service) into a zip
suggest-fixHigh-level remediation suggestions based on collected data

API examples

List available tools:

curl -H "x-api-key: changeme" http://localhost:3210/mcp/tools

Run a tool (JSON RPC-ish payload)

curl -X POST http://localhost:3210/mcp/run \
  -H "Content-Type: application/json" \
  -H "x-api-key: changeme" \
  -d '{"tool":"disk-usage","args":{"threshold_percent":80}}'

Typical JSON response

{
  "tool":"disk-usage",
  "status":"ok",
  "data":[{"mount":"/","used_percent":87,"recommendation":"clean /var/log or increase disk"}]
}

Use Cases

  • Desktop troubleshooting with an assistant: A user reports slow performance. The assistant calls sysinfo and processes tools to detect high CPU processes and responds with step-by-step recommendations (kill process, clear caches) and commands to run.
  • Pre-deployment checks: Before shipping a release to a developer machine, an automation agent runs disk-usage, network-check and service health tools to confirm readiness and returns a pass/fail report.
  • Remote helpdesk workflows: Support staff instruct an MCP-enabled assistant to collect logs and run port-scan to diagnose connectivity issues, then share a sanitized diagnostics bundle with engineering.
  • Incident triage automation: Integrate the MCP server into monitoring playbooks so the assistant can gather context (recent logs, service restarts) automatically and provide remediation or rollbacks.

Security and Best Practices

  • Run the MCP server only on hosts you control; limit API access with API keys and ALLOWED_ORIGINS.
  • Do not run as root unless necessary; prefer a dedicated low-privilege user.
  • Review the tools the server exposes—disable or harden any tool that can run arbitrary commands.
  • Rotate API keys and audit server logs regularly.

Resources

  • GitHub: https://github.com/vishnurudra-ai/KYC-mcp-server
  • MCP (Model Context Protocol): follow the MCP client documentation for your assistant (Claude Desktop or other MCP-capable clients) for registration and capabilities integration.

For development, extend the tools directory with new handlers that return structured JSON and text recommendations so AI agents can reason deterministically about host state. Contributions and issues are welcome on the GitHub repository.

Tags:developer-tools