# Node-RED MCP > Node-RED MCP lets you control complex, event-driven workflows directly through your AI agent. Manage entire IoT pipelines and low-code applications without leaving the chat window. You can check system diagnostics, update flow logic, or install new node modules just by asking your agent. ## Overview - **Category:** loved-by-devs - **Price:** Free - **Tags:** low-code, event-driven, workflow-automation, flow-management, system-diagnostics, api-orchestration ## Description Running automated systems means dealing with constantly changing flows, dependencies, and runtime errors. This MCP gives you a single connection point to manage those complexities in Node-RED. Instead of jumping between the editor GUI, terminal logs, and configuration files, you talk to your AI client. Your agent handles reading existing flow configurations, updating nodes, or checking system health. When you connect this capability via Vinkius, it becomes part of a larger catalog of tools, letting you orchestrate much more than just Node-RED. You can ask your agent to troubleshoot a memory spike and get the live diagnostics report instantly. This means complex, mission-critical automation is now controllable through simple conversation. ## Tools ### add_flow Creates a brand-new, empty workflow tab within your Node-RED instance. ### delete_flow Removes an entire flow configuration (tab) from the active workspace. ### get_diagnostics Retrieves a detailed report on system diagnostics, including OS version and memory usage. ### get_flow Pulls the configuration details for one specific workflow tab you want to inspect. ### get_flows Lists all currently active flow configurations available in your Node-RED environment. ### get_nodes Retrieves a list of every node module that is currently installed on the system. ### install_node Adds and installs a specified new node module package into your environment. ### remove_node Deletes an installed node module package from the system to clean up dependencies. ### set_flows Sets or switches the active flow configuration that Node-RED is currently focused on. ### get_settings Fetches all current runtime operational settings for the entire Node-RED instance. ### update_flow Modifies and saves changes to a specific, existing flow configuration tab. ## Prompt Examples **Prompt:** ``` Show me all active flows in my Node-RED instance. ``` **Response:** ``` I've retrieved your flows. You have 3 main tabs: 'Home Automation' (ID: f123), 'Data Logger' (ID: f456), and 'Alerts' (ID: f789). Would you like to inspect a specific one? ``` **Prompt:** ``` Check the system diagnostics and memory usage. ``` **Response:** ``` Diagnostics report: Node.js v18.16.0, OS: Linux. Memory usage is currently at 150MB of 512MB allocated. The system is healthy. ``` **Prompt:** ``` Install the 'node-red-node-email' module. ``` **Response:** ``` Starting installation of 'node-red-node-email'... The module has been successfully installed and is now available in your palette. ``` ## Capabilities ### Manage full flow configurations Create new workflows, delete unused tabs, retrieve existing flows, or modify running logic. ### Handle node dependencies List all installed modules, install missing packages, and remove obsolete nodes from your system. ### Check system health metrics Get immediate diagnostics on the Node.js version, operating system details, and current memory usage of the runtime environment. ### Retrieve runtime settings Fetch the active operational settings to understand what constraints or variables your workflow is running under. ## Use Cases ### The critical flow failure A smart home automation fails after a firmware update. Instead of logging into the server to check logs, you ask your agent to get_diagnostics. You see Node.js is running an outdated version and need to upgrade. Your agent reports the current setup (get_settings), allowing you to quickly determine the necessary patch. ### Adding a new data source You need your IoT dashboard to pull weather data, but the node isn't installed. You tell your agent, and it uses install_node. It confirms the installation, and then you use get_nodes to verify the module is now available for building the flow. ### Auditing an old project A new team member needs to understand a complex workflow built months ago. You ask your agent to get_flows, which lists all active tabs. You then use get_flow on specific IDs to show the entire structure without having to manually recreate or navigate the whole thing. ### Refactoring an obsolete module A project is winding down and you need to clean up dependencies. Instead of searching your file system, you use get_nodes to list everything installed. You then instruct the agent to remove_node for outdated modules. ## Benefits - You check the overall state of your environment using get_diagnostics. You instantly know if a memory spike is due to OS limits or an internal application leak, without manually checking server logs. - Updating complex logic used to require multiple manual steps: copying JSON definitions and pasting them into different dashboards. Now you can use update_flow and tell your agent exactly what change needs implementing. - Dependency management gets simple. Instead of remembering the npm command or searching documentation, you just ask your agent to install a missing piece using install_node. - The ability to see everything with get_flows is huge. You immediately know which automation pathways exist and can use get_flow on any specific tab without navigating away from your chat interface. - You manage the entire lifecycle of an application: list dependencies with get_nodes, add new ones with install_node, and remove them later with remove_node—all conversationally controlled. ## How It Works The bottom line is that your AI client routes complex flow and system commands through this MCP, giving you full control over the entire Node-RED environment from a single chat interface. 1. First, subscribe to this MCP and provide your Node-RED Base URL and Access Token. 2. Next, authorize your preferred AI client—Claude, Cursor, or any other MCP-compatible agent—to access the connection. 3. Finally, tell your agent what you need. You can ask it to 'show me all active flows' or 'install the weather node,' and it executes the command instantly. ## Frequently Asked Questions **How do I check if my Node-RED instance has enough memory using the Node-RED MCP?** You run get_diagnostics. This tool provides a full report, including live details on OS and current memory usage, letting you know immediately if resources are constrained. **Can I see all my existing workflows using the Node-RED MCP?** Yes, use get_flows. This tool lists every active flow configuration available in your workspace, giving you a clear overview of your automation paths. **What if I need to add a new node module? Which tool should I use with the Node-RED MCP?** You use install_node. Simply tell your agent what package name needs adding, and it handles the installation process for you. **How do I update an old flow without losing data using the Node-RED MCP?** You can use update_flow. You provide the specific ID of the flow you want to change, and your agent applies the modifications while keeping the existing structure intact. **Does the Node-RED MCP let me see what settings my instance is running?** Yes, run get_settings. This tool fetches all current runtime operational parameters for the entire system, so you know exactly how your application behaves out of the box.