Subscribe to our newsletter
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
.png)
From dashboard data to actionable workflows
Data is the key to running an efficient business today. But how much data do you and your team operate with every day? How many dashboards, reports, alerts, and metrics do you need to check before acting? For Kubernetes teams, this problem is more than real: cost reports, utilization metrics, workload risks, right-sizing recommendations, deployment history, incident context, and logs. The challenge is not getting more data. The challenge is cutting through the noise and getting the right insight at the right time.
If you are a PerfectScale user, PerfectScale already surfaces the data that matters. But until now, this data was available mainly through the PerfectScale dashboard, which creates some workflow friction.
“When teams need to export data, build custom reports, compare metrics from different sources, or ask an AI agent to investigate anomalies, summarize findings, or explain risks, a UI-only model becomes an imitation," said Amit Bezalel, Chief Solution Architect of PerfectScale by DoiT. "The future of infrastructure management is moving toward agentic AI, and that future is already taking shape. For DevOps, Platform, and SRE teams, optimization data needs to be available where work already happens: in terminals, scripts, internal tools, and AI-powered workflows.”
The challenge is clear, and the solution is simple. That is how PerfectScale AI agent skill was born, bringing PerfectScale data into agentic workflows.
AI-powered workflows with PerfectScale data
This skill enables AI-powered coding and operations agents, such as Claude Code, OpenAI Agents SDK, and others, to work with PerfectScale data. Instead of opening dashboards, switching between tools, or writing queries, engineering teams can simply ask a question and let the agent find the answer.
PerfectScale CLI gives agents the data they actually need, making Kubernetes optimization data easy to query, filter, parse, and combine with other operational data sources:
- Structured output: JSON and JSONL for seamless parsing.
- Rich filtering: Agents can retrieve results by cluster, namespace, workload type, cost, risk, and more.
- Composable data: Agents can combine PerfectScale data with other sources to create insights that no single tool can provide on its own.
What this looks like in practice
Once the skill is installed, ask Claude or another agent you use a question such as: “What are the top 10 most wasteful workloads in aws-us-prod?”. In seconds, the agent will return a clear answer with dollar amounts and any other requested information.
With flexible agent access to Kubernetes optimization data, teams can easily list and inspect clusters, analyze workloads by cost, waste, or risk, group results by namespace, workload type, policy, or label, and more, simplifying Kubernetes operations across existing workflows and improving optimization results.
That is the shift: PerfectScale data is no longer just something you open in a dashboard. It is something your tools and agents can use directly.
Getting started
PerfectScale CLI is open source and available now. Download it from GitHub, generate an API token in your PerfectScale dashboard, and run your first requests within two minutes. Watch the short video below, where Anton Weiss walks you through the entire process, from installation to insights.
📖 Full documentation is available at the PerfectScale docs portal.
What is next
This is just the beginning. We’re adding support for configurable time periods, server-side filtering, nodegroup commands, and more export formats.
The future of infrastructure management is moving toward agentic AI. PerfectScale CLI supports this shift by making Kubernetes optimization data easier to access, combine, and act on across the tools and workflows teams already use.
Not using PerfectScale yet? Start for free today and see it yourself.
