# Netdata MCP > Netdata MCP connects your entire infrastructure monitoring suite to any AI client. Instantly pull real-time performance metrics, check current system alerts, and get deep operational data for specific nodes or across entire cloud spaces. Your agent reads the raw numbers so you don't have to click through dozens of dashboards. ## Overview - **Category:** cloud-infrastructure - **Price:** Free - **Tags:** real-time-monitoring, infrastructure-observability, system-metrics, performance-analysis, alert-management ## Description Your AI client can connect directly to Netdata to give you instant visibility into your infrastructure's performance. Instead of jumping between monitoring tabs, you ask a question—like 'Why is CPU spiking on Node Alpha?'—and get an immediate answer backed by real data. You can fetch granular metrics for specific components like RAM or network throughput using one command. Need to know if something is broken? Your agent checks active alarms across your local machines or even monitors critical issues space-wide. When you're ready to analyze the raw numbers, you retrieve all collected metrics in a format that external tools love. This makes troubleshooting faster and less painful. Through Vinkius, you connect this powerful MCP with any compatible client, giving your agent 24/7 System Administrator capabilities right where you work. ## Tools ### get_alarms Retrieves the current status of every configured alarm on a machine. ### get_all_metrics Gathers all available performance metrics into one bulk dataset for scraping or analysis. ### list_charts Provides a list of every metric chart currently tracked on the connected node. ### get_chart_data Fetches specific time-series data points from any requested performance chart. ### get_agent_info Returns detailed information, including the version and host details, for the Netdata Agent. ### list_room_nodes Lists all individual nodes belonging to a specific monitored room. ### list_rooms Provides a list of all rooms within a defined monitoring space. ### list_space_alerts Aggregates and reports on critical issues across the entire monitored cloud space. ### list_space_nodes Lists all machines connected under a specific monitoring space. ### list_spaces Retrieves a comprehensive list of all available Netdata Cloud spaces. ## Prompt Examples **Prompt:** ``` Get the current information and version of the Netdata agent. ``` **Response:** ``` I've retrieved the agent info. The host is 'prod-server-01' running Netdata version v1.44.1 with 128 enabled collectors. ``` **Prompt:** ``` List all available charts on this node so I can see what metrics are being collected. ``` **Response:** ``` I've compiled the list of charts. You have access to metrics for 'system.cpu', 'system.ram', 'net.eth0', and 45 other dimensions. Which one would you like to analyze? ``` **Prompt:** ``` Are there any active alarms or warnings on the local agent right now? ``` **Response:** ``` Checking alarms... I found one warning for 'disk_backlog' on /dev/sda. All other system health checks are currently clear. ``` ## Capabilities ### Analyze system performance data Fetch detailed metric readings—like CPU load or disk usage—for a specific component on a machine. ### Manage current alerts Check for active warnings and alarms, either on a single local agent or across an entire cloud environment. ### View infrastructure inventory List all connected spaces, rooms, and individual nodes to understand your full monitoring scope. ### Export raw metrics data Collect every available metric across the monitored nodes into a single dataset for external processing. ## Use Cases ### Investigating a sudden network slowdown The user asks their agent about recent network issues. The agent checks `list_space_nodes` to identify all relevant machines, then uses `get_chart_data` on the network metrics for those nodes. It returns specific data confirming which machine is spiking and why. ### Performing a routine system audit The user needs a full picture of the environment. The agent first calls `list_spaces` to see all environments, then uses `list_rooms` to drill down into each one, building a complete map of connectivity. ### Responding to an urgent production alarm A critical alert hits. The user prompts the agent for active alarms using `get_alarms`. The agent immediately returns which specific component is failing and whether other system health checks are clear, saving minutes of manual investigation. ### Preparing data for a quarterly report The team needs historical performance metrics. Instead of exporting manually from dashboards, the user runs `get_all_metrics`, getting all raw data in one go that can be fed directly into reporting tools. ## Benefits - Instant bottleneck diagnosis: Instead of guessing, ask your agent to fetch metric data from a specific chart using `get_chart_data` to pinpoint exactly where performance is dipping. - Centralized alert visibility: Check for local machine issues with `get_alarms`, or get a high-level view of critical alerts across the whole space using `list_space_alerts`. - Full inventory mapping: Never lose track of your assets. Use `list_spaces` and `list_rooms` to map out every node connected through your infrastructure. - Historical data analysis: Need to feed metrics into a custom tool? Run `get_all_metrics` to pull all raw performance numbers for deep external processing. - Quick health checks: Before starting an investigation, run `get_agent_info` to confirm the version and basic status of the agent itself. ## How It Works The bottom line is you get immediate operational insights without ever leaving your IDE or terminal. 1. Subscribe to this MCP and provide your Netdata Cloud Token or Agent URL. 2. Your AI client accesses the connection details, giving it visibility into all connected nodes and spaces. 3. You prompt your agent with a question (e.g., 'What are the current disk alerts?') and receive specific, actionable data. ## Frequently Asked Questions **How do I check all nodes connected to my environment using Netdata MCP?** You can list all spaces first using `list_spaces`. Then, drill down into the rooms and finally use `list_space_nodes` to get a complete inventory of every machine monitored. **Can I check for active alarms with Netdata MCP?** Yes. You can run `get_alarms` to see local warnings, or you can use `list_space_alerts` if you need a summary of critical issues across the entire cloud space. **What data do I get when I use get_chart_data with Netdata MCP?** You get specific, time-series metric readings for any chart type, like CPU or RAM. This allows you to diagnose performance bottlenecks precisely rather than just seeing a general graph. **Does Netdata MCP help me analyze historical data?** Yes, you can use `get_all_metrics` to retrieve all collected metrics in a format designed for external analysis tools, making history accessible. **Is Netdata MCP just for Linux servers?** The MCP is designed to work with your configured Netdata Agent. It accesses the data available through that specific agent connection.