One MCP server. Every pillar of your cluster: cost, security, reliability, policy. Ask the questions your team actually asks. Get answers that cross product boundaries, not another dashboard to babysit.
Insigh8s sits between your AI assistant and your Kubernetes stack. Each team asks the question they care about. The MCP fans out to the right sources, joins the data, and returns one answer with remediation guidance.
Each tool answers one clear question. An SRE investigating a problem, a FinOps engineer tracking spend, and a compliance reviewer running an audit all want different answers, so they get different tools. No god-tool that tries to do everything.
Anyone can call an API. The hard part is knowing what to look at, in what order, with what thresholds, and which finding actually matters. Insigh8s encodes that judgement into every tool.
investigate_namespace finds unhealthy pods, it also shows the deploy that caused them and the error pattern in logs.
list_audit_frameworks().
kubectl commands. Ship the fix, not the homework.
With a pile of raw MCPs, the AI has to guess which tools to call, how to stitch the results, and what matters. That guess changes every time. With Insigh8s, the orchestration lives inside a tested tool, so the answer is the same whether you're on Claude, GPT, Gemini, or a local Llama.
audit_namespace v1.2 is a reviewable diff. Your security team can read what it checks.
Every team has its own dashboard. Every dashboard answers one question. And when something breaks at 2am, you're still the correlation engine.
You could install a dozen MCP servers. But raw data isn't triage. Answers are.
# you, manually: kubectl_get_pods("payments") → which pods are failing? kubectl_describe("api-gw-...") → OOMKilled, but why now? kubectl_rollout_history(...) → 3 recent deploys, which one? prometheus_query("rate(...)") # you're writing PromQL at 2am.
# your AI, once: investigate_namespace( namespace="payments", window="15m" ) ↳ api-gateway OOMKilled x3 ↳ correlates to deploy d4c7f19 (14:32 UTC) ↳ error rate: 3% → 47% ↳ likely cause + rollback # ~6 seconds.
Your AI can already call kubectl. What it can't do is tell you which 40 pods to look at first. Here's what we're building, and why composite tools beat raw MCPs.
Read post →A walkthrough of what a single SRE triage call actually does under the hood: the data sources, the join logic, and the opinions encoded in Go.
Follow on Hashnode →How Insigh8s talks directly to the underlying CNCF tools (not their MCPs) and what happens when one of them isn't installed on the cluster.
Follow on Hashnode →Insigh8s is a community project, Apache 2.0 licensed. The design is open, the roadmap is public, and the first release is taking shape. If the composite-tool approach resonates with you, there are a few ways to get involved.
Following the repo is the best way to track progress, see the roadmap take shape, and be notified when v0.1 drops. No commitment.
GitHub ↗What composite tools would you find useful? What thresholds should be defaults? The early decisions are still open. Join the GitHub Discussions and help set direction.
Discussions ↗Good-first-issue labels, a simple architecture, and a roadmap you can pick from. Platform engineers, FinOps folks, security people, and writers all welcome.
Contribute ↗One email. No marketing. Sent when the first release lands and is stable enough to install.