Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    System Stats — Local Performance Diagnostics

    1

    Live hardware diagnostics and performance monitoring for CPU, RAM, GPU, and disk metrics via MCP.

    Free

    1 useSecurity scanned
    System Stats — Local Performance Diagnostics

    System Stats — Local Performance Diagnostics

    Example session with this skill installed

    My computer feels slow. Use System Stats to compare CPU and RAM across several samples, identify resource-heavy applications, and explain your findings and any measurement limitations. Do not change system settings or close applications.

    • Read your context and instructions
    • Compiled the system stats

    Illustrative example

    The following measurements are fictional and demonstrate the report format.

    Finding

    RAM usage remained above 90% throughout the short sampling window. This may contribute to the slowdown, but the measurements do not establish the complete cause.

    Evidence

    MeasurementResult
    Sampling windowThree samples over 2.6 seconds
    CPU utilization18%, 25%, 20% — average 21%
    RAM utilization92%, 93%, 92% — average 92.3%
    Largest observed memory consumerBrowser processes, approximately 5.8 GiB
    CPU temperatureUnavailable

    Interpretation

    RAM pressure persisted across all three samples. CPU utilization did not show sustained pressure during this window. The browser was the largest observed memory consumer, making it a candidate for further investigation.

    Limitations

    The missing temperature reading cannot be interpreted as a normal temperature. A short sampling window cannot establish a memory leak, and the process ranking alone does not prove causation. Disk latency was not measured.

    Next steps

    1. Save your work and review unnecessary browser tabs or active tasks.
    2. Repeat the diagnostic while the slowdown is occurring.
    3. Compare memory usage before and after an application is closed manually.

    No applications were terminated, files deleted, or system settings changed.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Give your AI agent a practical workflow for diagnosing local computer performance with mcp-system-stats.

    What it does

    • Compares CPU and RAM across multiple samples to distinguish isolated spikes from sustained pressure.
    • Identifies resource-heavy applications and processes.
    • Examines disk capacity, GPU memory, and network activity when relevant.
    • Explains missing sensors, unsupported features, and measurement limitations.
    • Produces clear findings and actionable recommendations based on real tool results.

    Requirements

    Requires the @fran0808/system-stats MCP server configured locally and an agent that can load skill instructions and access local stdio MCP tools. Loading this skill alone does not connect the server.

    Setup instructions cover npm installation, an npx alternative, and separate configuration examples for Cursor, Claude Desktop, Claude Code, and Codex.

    Included

    Diagnostic instructions, connection guidance, configuration templates, tool limitations, and example reports.

    Limitations

    Telemetry availability depends on the operating system, hardware, drivers, and permissions. The server measures the machine where it runs. This skill provides read-only diagnostics; it does not automatically close applications, delete files, or change system settings.

    Free and open source under the MIT license.

    Requires local mcp-system-stats server via stdio. Service/startup inventory is OS-specific (mostly Windows). Cannot terminate processes or modify system settings.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 29 days ago

    • Free with an Agensi account

    Requires

    Node.js
    without
    macOS
    alternative.
    or
    substitute

    Permissions

    Write Files
    Network Access
    Listed29 days ago

    What's inside

    Frequently Asked Questions

    This skill provides real-time telemetry—such as CPU load, VRAM usage, and disk health—allowing your AI agent to identify the specific root cause of a slowdown instead of offering generic troubleshooting advice.

    Yes, it is fully compatible with Windows, Linux, and macOS, though certain advanced features like startup program management are currently focused on Windows and Linux environments.

    The skill requires an environment that supports the Model Context Protocol (MCP) via stdio, such as Claude Desktop or other MCP-compatible agent hosts.

    Absolutely—the skill includes specialized GPU monitoring tools to track VRAM utilization and compute processes, which is essential for managing local LLM workloads.

    No, this skill is designed for local machine diagnostics only and does not support monitoring remote servers or cloud infrastructure.

    You receive the standardized MCP tool definitions and instructions required to connect your agent to the mcp-system-stats server (v2.0.1+), enabling immediate hardware data access.

    New in LLM Ops