# Flystack MCP MCP > Flystack MCP connects your agent directly to your cloud infrastructure data. You can instantly audit deployments, check cluster health, and pull real-time performance metrics without ever touching a console or running complex CLI commands. ## Overview - **Category:** cloud-infrastructure - **Price:** Free - **Tags:** infrastructure-auditing, cluster-management, deployment-monitoring, performance-metrics, cloud-ops ## Description Your agent acts like an internal DevOps consultant for your entire stack. Instead of logging into separate cloud dashboards to check status codes or gather usage numbers, you just ask it. It pulls together the operational truth—active deployment statuses, cluster configurations, and CPU/memory metrics—and presents it in plain language. Whether you're running a quick audit on staging environments or need detailed resource markers for a production issue, your agent gets you the up-to-the-minute data instantly. You just connect Flystack via Vinkius to your preferred AI client and treat complex infrastructure management like a natural conversation. ## Tools ### check_api_status Determines if the entire Flystack monitoring system is currently operational and available for use. ### get_cluster_metrics Pulls real-time performance data, like CPU usage and memory consumption, for a specific infrastructure cluster ID. ### get_deployment_details Retrieves comprehensive metadata about a single service deployment using its unique ID. ### list_flystack_clusters Generates a complete list of all infrastructure clusters managed within your Flystack account. ### list_flystack_deployments Retrieves an inventory of every currently active deployment across your entire cloud environment. ## Prompt Examples **Prompt:** ``` List all active deployments using Flystack. ``` **Response:** ``` I've retrieved the list of deployments! You have 3 active services, including 'web-production' and 'api-gateway'. Would you like the full details or performance metrics for any of these? ``` **Prompt:** ``` What are the metrics for cluster ID 'cluster-123'? ``` **Response:** ``` I've retrieved the real-time metrics for cluster-123! The CPU usage is currently 45% and memory is at 60%. I can assist you with an audit of specific deployment IDs within this cluster if you'd like. ``` **Prompt:** ``` Show details for deployment 'api-gateway'. ``` **Response:** ``` I've identified the deployment 'api-gateway'! It is running in the 'us-east-1' region and was last updated 2 days ago. I can provide the unique ID and cluster metadata to help you identify it in your workflow. ``` ## Capabilities ### Audit deployment status Retrieves high-resolution metadata for any active service, showing its current region and creation details. ### Check cluster health Runs an audit on your infrastructure clusters to verify their operational configuration and state. ### Monitor resource usage Queries real-time performance data, including current CPU utilization and memory consumption for a given group of services. ## Use Cases ### Investigating production resource spikes A developer notices high latency and asks their agent: 'What are the current metrics for cluster ID X?' The agent uses `get_cluster_metrics` to immediately report 95% CPU usage, allowing the team to scale resources before service degradation occurs. ### Pre-launch compliance audit A CISO needs proof of deployment status across all regions. They ask their agent to run `list_flystack_deployments` and check each result for required metadata, confirming every service is active. ### Troubleshooting a failing service An engineer suspects an old deployment version is causing issues. They use `get_deployment_details` on the specific ID to confirm its creation date and region metadata before rolling back the change. ### Initial infrastructure discovery A new team member needs a full understanding of the environment. They ask their agent to run `list_flystack_clusters` first, getting an immediate inventory of all monitored operational groups. ## Benefits - You stop jumping between multiple consoles. Instead, your agent consolidates deployment status and cluster metrics into one conversation thread. - It eliminates the need for manual data collection. With tools like `list_flystack_deployments`, you get an immediate inventory of every active service. - Get deep resource visibility on demand. Use `get_cluster_metrics` to pull real-time CPU and memory usage without running specialized monitoring scripts. - Deep-dive into specific services using `get_deployment_details`. This gives you the unique ID, status, and region metadata for any single service instance. - `list_flystack_clusters` provides a clear map of your entire infrastructure footprint, letting you scope audits quickly before asking about metrics. ## How It Works The bottom line is, you get structured operational data without writing any configuration files or shell scripts. 1. Subscribe to the Flystack MCP via Vinkius and provide your API Key. 2. Ask your AI agent (via Claude, Cursor, etc.) to perform a specific audit or check metrics. 3. The MCP executes the query against the infrastructure data sources and returns the required status information. ## Frequently Asked Questions **How do I check if Flystack MCP is working with `check_api_status`?** Running `check_api_status` confirms the entire monitoring service is operational. If it returns a success code, you know your agent can proceed safely to query metrics or deployments. **What does `list_flystack_deployments` return?** This tool gives you an inventory of every active deployment across your account. You get the list needed to then select specific IDs for further detail checking. **Can I check cluster metrics for a whole group? How does `get_cluster_metrics` work?** You need to specify a single cluster ID when using `get_cluster_metrics`. It then returns the real-time CPU and memory usage numbers you requested. **Do I need an API key for Flystack MCP? What about deployment IDs?** Yes, you must supply your API Key to connect. When asking for details using `get_deployment_details`, you'll need the specific deployment ID. **If I need to audit several environments, how do I chain `list_flystack_clusters` with `get_cluster_metrics`?** The tool returns a list of available cluster IDs. You must pass these identifiers sequentially or instruct your agent to iterate through the full list to gather metrics for every environment. **When I run `get_deployment_details`, what specific metadata does it provide about the deployment?** It provides high-resolution data, including the deployment's unique ID, its physical region, and its original creation metadata. This is key for deep infrastructure classification. **What happens if I try to use `get_deployment_details` with an invalid or deleted deployment ID?** The MCP handles this gracefully by returning a specific error code and clear message stating that the resource doesn't exist. Your agent can then report the failure without stopping the workflow. **If I have many services, are there any limits on how much data `list_flystack_deployments` provides?** The tool lists all currently active deployments linked to your account. For extremely large inventories, you may need to implement pagination logic within your agent's code to pull all records.