A Semantic Kernel plugin to diagnose and monitor system health on a Windows machine. This plugin exposes various system-level diagnostics as functions, allowing LLM agents to query the state of the machine using natural language.

Get the latest pre-compiled version from the Releases page.

1. Download the latest WorkstationInsights.zip.
2. Extract to a folder of your choice.
3. Run WorkstationInsights.exe.

• OS and system info
• CPU, memory, and disk usage
• Running processes and services
• Event logs and uptime
• Network and ping tests
• Environment variables

The core of this plugin is an open-ended PowerShell execution engine that allows the AI to craft and execute any diagnostic script.

• Execution Guardrails: Timeouts and safety prompts prevent hung or dangerous operations.
• Structured Data: The AI favors ConvertTo-Json output for highly accurate diagnostics.
• Auto-Truncation: Intelligent handling of large command outputs to stay within LLM context limits.

"Why is my computer lagging?"

• Calls GetCpuUsage, GetMemoryUsagePercent, ListTopCpuProcesses, etc.
• LLM provides a diagnosis like: "Your CPU usage is at 85%, and Chrome is using the most resources."

Daily or hourly health scans

• Calls uptime, updates, disk info, and running services
• Summarized into a status report by the LLM

"Is the internet down?"

• Calls PingTest("8.8.8.8"), GetActiveNetworkInterfaces
• LLM summarizes connection status and suggestions

"Why won’t SQL Server start?"

• Calls CheckServiceStatus, GetEventLogs("Application")
• LLM diagnoses from service state and logs

"How much RAM and CPU does this machine have?"

• Calls GetTotalPhysicalMemory, GetLogicalProcessorCount
• LLM replies with a concise system spec

Alert when RAM > 90%

• Kernel auto-checks memory and CPU
• Triggers alert or flow when thresholds are exceeded

"How can I free up space?"

• Uses GetDriveInfo, ListTopMemoryProcesses
• GPT suggests uninstalling or clearing large files

1. Add this plugin class to your Semantic Kernel project.
2. Register with the Kernel:

builder.Plugins.AddFromType<WorkstationInsights>();

3. Now your LLM can invoke diagnostics like:

"Show the top memory-using processes"

This plugin uses:

• System.Diagnostics
• System.Management
• System.IO
• System.ServiceProcess
• System.Net.NetworkInformation

No external dependencies are required.

MIT — use freely in personal or commercial Semantic Kernel applications.