Infrastructure mcp operator

    1

    Securely operate and inspect servers, clusters, and network devices via infrastructure-mcp.

    Free

    1 useSecurity scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Infrastructure mcp operator

    Example session with this skill installed

    I'm seeing high latency on the production web-tier. Use the infrastructure-mcp tools to list the servers in the 'web-prod' group and check their current CPU load.

    • Read your context and instructions
    • Compiled the infrastructure mcp operator

    I have queried the infrastructure-mcp server to inspect the 'web-prod' inventory group. I retrieved the status for three active nodes and checked their resource utilization.

    Infrastructure Status: web-prod group

    Server IDHostnameIP AddressCPU UsageMemory UsageStatus
    srv-01-wpweb-prod-0110.0.4.1288.4%4.2GB / 8GBHealthy
    srv-02-wpweb-prod-0210.0.4.1392.1%5.8GB / 8GBThrottled
    srv-03-wpweb-prod-0310.0.4.1445.0%2.1GB / 8GBHealthy

    Assessment

    High Load: web-prod-02 is currently throttled with CPU usage exceeding 90%. web-prod-01 is also near capacity.

    Imbalance: web-prod-03 is under-utilized at 45%, suggesting a possible load balancer distribution issue or a specific process hang on the other nodes.

    • Evidence: Timestamps from list_servers indicate these metrics were captured within the last 30 seconds.

    Unknowns

    • The specific process causing the spike on web-prod-02 is unknown as detailed process monitoring tools were not invoked.
    • Load balancer logs are not included in this view.

    Next steps

    1. Check the load balancer distribution logs for the web-prod pool.
    2. Inspect the top processes on web-prod-02 using the get_server_metrics detail view.
    3. Verify if any recent deployments or cron jobs started at the time of the latency spike.

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

    About this skill

    Managing complex infrastructure usually requires juggling multiple CLI tools, credentials, and context windows. This skill operates infrastructure via an existing MCP server, providing a secure bridge to inspect servers, containers, and network devices without manual credential handling.

    What it does

    • Discover targets by listing servers and tags to identify the correct environment.
    • Inspect resources across clusters, hypervisors, and Cisco devices using named tools.
    • Collect evidence by pulling real-time status and logs from infrastructure components.
    • Analyze health by interpreting tool outputs to identify specific points of failure.
    • Execute changes using a controlled flow that requires explicit user approval for mutations.

    How it works

    1. Identify domain by mapping the user request to specific infrastructure targets.
    2. Verify tools by checking the available MCP surface for the required adapters.
    3. Gather data starting with broad status checks before drilling into specific logs.
    4. Report findings separating raw tool observations from logical interpretations.

    Frameworks & tools

    This skill requires a configured infrastructure-mcp server. It interacts with environments involving Docker, Kubernetes, Cisco networking, and various hypervisors through standardized tool interfaces.

    Why this beats prompting it yourself

    Generic prompts often hallucinate resource names or suggest unsafe execution commands. This skill enforces a strict safety contract, preventing the use of unrestricted shell escapes and ensuring every mutation is explicitly confirmed against real inventory names.

    Use cases

    • Troubleshoot container crashes in a Kubernetes cluster without manual auth.
    • Audit Cisco device configurations across multiple network segments.
    • Verify hypervisor resource allocation during performance degradation.
    • Perform read-only fleet health checks before scheduled maintenance.

    Known limitations

    Requires an active infrastructure-mcp server installation. It cannot use unrestricted execution tools like run_command or kubectl_exec for security reasons.

    Requires Infrastructure MCP Server. This skill does not include or host any infrastructure access. You must deploy and connect Infrastructure MCP Server (free, MIT, self-hosted) first. Without it, the skill will tell you it can't run.

    Setup (about 10 minutes):

    1. Install the infrastructure-mcp server (Docker, Docker Compose, or from source). https://github.com/radumalica/infrastructure-mcp
    2. Add your servers to the inventory file. Credentials stay on your server.
    3. Connect your agent over stdio or authenticated HTTP.
    4. Install this skill.

    Not included: hosted service, infrastructure access, credentials, API credits, or support. Works with Linux, Docker, Kubernetes, Grafana, Proxmox and Cisco today.

    Requires a pre-configured infrastructure-mcp server. Restricted from using unrestricted execution escapes like docker_exec or run_command.

    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 1 month ago

    • Free with an Agensi account

    Requires

    Infrastructure MCP

    Permissions

    No special permissions declared or detected

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    This skill provides a secure interface to manage servers, Kubernetes clusters, and Cisco devices through a standardized MCP server, eliminating the need to manually juggle multiple CLI tools and credentials.

    No, this skill is designed with a 'safety-first' architecture that prevents unrestricted shell access and requires explicit user confirmation before any state-changing actions are executed.

    You must have an active `infrastructure-mcp` server installed and configured; the skill acts as the intelligent bridge between your AI agent and that specific server interface.

    The skill supports discovery and inspection for Docker containers, Kubernetes clusters, various hypervisors, and Cisco networking hardware via supported MCP adapters.

    By using real-time inventory discovery rather than relying on the LLM's memory, the skill ensures commands are only sent to verified, existing resource names, preventing errors common in generic prompting.

    You receive the skill definition and optimized logic for your AI agent; please note that the underlying `infrastructure-mcp` server hardware or software must be provided by your own DevOps environment.

    New in DevOps & Cloud