AI-Powered Proxmox Management Agent - n8n Workflow

• System administrators and DevOps engineers managing Proxmox VE environments.


• Users looking for advanced n8n templates combining AI and infrastructure management.


• Developers needing to execute complex Proxmox API calls via simple text commands.


• Anyone utilizing the n8n platform for advanced automation tasks.

Managing virtualization environments like Proxmox often requires manual interaction with complex APIs or web interfaces. This sophisticated n8n workflow solves this challenge by implementing a custom AI Agent, making Proxmox management conversational. The agent, powered by the Google Gemini language model, is specifically trained using the Proxmox API documentation and wiki to securely convert user requests (e.g., 'Create a VM with ID 201,' 'List all VMs on psb1') into precise, structured API calls. The core value of this n8n template lies in its ability to securely authenticate and execute complex infrastructure changes based on natural language, dramatically simplifying DevOps operations and showcasing the power of the n8n node structure for infrastructure-as-code automation.

This n8n workflow operates through a multi-step conversational process starting with an input trigger and culminating in a structured response from the AI. The flow begins with an incoming user request captured by an n8n trigger (such as When chat message received, Telegram Trigger, or Gmail Trigger).

1. Input & Context: The user prompt feeds into the main AI Agent n8n node.


2. Tooling: The AI Agent is equipped with three crucial tools: Proxmox API Documentation, Proxmox API Wiki, and a live Proxmox API endpoint connection. This context allows the agent to formulate highly accurate, structured API requests.


3. Structured Generation: The AI agent generates a JSON output containing the desired responsetype (HTTP method: GET, POST, DELETE, PUT) and the target url and details payload. This output is enforced by an Auto-fixing Output Parser and a Structured Output Parser n8n node, ensuring data integrity.


4. Execution Routing: A Switch n8n node routes the flow based on the determined responsetype (e.g., 'GET' goes to HTTP Request, 'POST' goes to If node, 'DELETE' goes to If1 node).


5. API Execution: The appropriate HTTP Request n8n node executes the call against the Proxmox server, using shared header authentication.


6. Response Handling (GET): For informational requests (GET), the raw JSON response is processed by a Code n8n node (Structure Response) and then summarized into human-readable text by a second AI model (AI Agent1/Gemini).


7. Response Handling (POST/DELETE): For actions (POST, DELETE), the resulting UPID response is processed by specialized Code n8n nodes (Structgure Response from Proxmox, Format Response and Hide Sensitive Data) to generate a clear confirmation message for the user, completing the n8n workflow cycle.

To deploy this powerful n8n workflow, follow these setup steps, focusing especially on the Proxmox credentials:

1. Import: Import this JSON file into your n8n instance using the 'New' menu and selecting 'Import from JSON'.


2. Proxmox API Key Setup: Proxmox requires specific token credentials. In your Proxmox Data Center, create an API Key. In n8n, create a new 'Header Auth' credential named 'Proxmox'.

1. Google Gemini Credentials: Create or select your existing 'Google Gemini (PaLM) Api account' credential and link it to the Google Gemini Chat Model n8n nodes.


2. Tool Configuration: Review the Proxmox tool n8n node parameters. Update the base URL (https://10.11.12.101:8006/api2/) and the node list (psb1, psb2, psb3) to match your specific Proxmox cluster configuration.


3. Trigger Setup: Configure your preferred n8n trigger (e.g., When chat message received or Telegram Trigger) and link its output to the main AI Agent n8n node to begin using this n8n templates solution.

This n8n workflow utilizes several specialized and core n8n node types to achieve complex automation:

AI Agent (1492e53e): The central decision-making n8n node. It uses the ReAct agent mechanism, guided by an extensive prompt defining rules, structure, and examples (e.g., handling missing parameters, defaulting to psb1). It translates conversational text into precise API calls.
Proxmox, Proxmox API Documentation, Proxmox API Wiki (ToolHttpRequest): These specialized n8n node tools provide the AI agent with external knowledge needed to construct accurate Proxmox API commands.
Structured Output Parser (8565ac2f): Ensures the AI's output conforms strictly to the expected JSON schema (containing responsetype, url, details), which is critical for subsequent steps.
Switch (9253d036): A flow control n8n node that examines the responsetype (GET, POST, DELETE, PUT) generated by the AI agent and directs the execution path to the correct HTTP method handler.
HTTP Request (e.g., 09444fa1, e6d85380): Multiple n8n node instances execute the final API calls against the Proxmox server, handling GET, POST, and DELETE methods using the dynamic URL and payload generated by the AI.
Code (f11ac59e & 60234199): Custom script n8n node instances used for two main purposes: 1) Structuring complex JSON responses into a readable format for the summarizing AI, and 2) Parsing the cryptic Proxmox UPID response strings (from POST/DELETE actions) into clear confirmation messages.