Skill Detail

Investigate Kubernetes network incidents through MCP with Kubeshark

Query live and historical Kubernetes network traffic through Kubeshark's MCP server when an agent needs packet-level evidence, API payloads, or service-path traces for incident response.

Monitoring & AlertsMCP
Monitoring & Alerts MCP Security Reviewed
⭐ 11.9k GitHub stars
INSTALL WITH ANY AGENT
npx skills add agentskillexchange/skills --skill investigate-kubernetes-network-incidents-through-mcp-with-kubeshark Copy
Works best when you want a reusable capability, not another fragile one-off prompt.
At a glance
Tools required
Kubeshark deployed in the target Kubernetes cluster and an MCP-compatible client
Install & setup
Install Kubeshark in the cluster with Helm (for example: helm repo add kubeshark https://helm.kubeshark.com && helm install kubeshark kubeshark/kubeshark –set mcp.enabled=true –set mcp.port=8898), then connect an MCP client to the Kubeshark MCP endpoint.
Author
Kubeshark
Publisher
Organization
Last updated
Apr 20, 2026
Quick brief

Use Kubeshark when an agent needs direct access to Kubernetes network evidence instead of guessing from logs alone. Kubeshark captures and indexes L4 and L7 traffic with Kubernetes context, exposes that data through MCP, and lets the agent ask focused questions about failed requests, service-to-service paths, TLS traffic, and retrospective packet captures.

How it works

What this skill actually does

Invoke it during root cause analysis, degraded API investigations, and cross-service incident response, especially when the important signal lives in the network path rather than in one workload’s logs. The scope boundary is tight: this skill is about querying captured cluster traffic and exporting targeted evidence through Kubeshark’s MCP tools, not about being a general observability platform card.