mcpField
K
kubesense-skills

kubesense-skills

github.com/kubesense-aiCloud & DevOps383 installsUpdated 3d ago
Repository

About kubesense-skills

kubesense-skills is a collection of observability skills for KubeSense, organized around telemetry surfaces. It guides agents through logs, traces, metrics, infrastructure, alerts, dashboards, SLOs, and query conventions, with most skills using the KubeSense MCP server.

Key features

  1. 01Log search and aggregation
  2. 02Trace latency and error analysis
  3. 03Metrics and infrastructure queries
  4. 04Alert and dashboard workflows
  5. 05SLO and Datadog migration guidance
  6. 06SQL and SPL query patterns

Use cases

  1. 01Investigating log errors and slow requests
  2. 02Reviewing Kubernetes or cloud metrics
  3. 03Creating alerts, dashboards, and SLOs
  4. 04Routing observability questions to relevant skills

MCP support

Most skills require a connected KubeSense MCP server for querying and validation. kubesense-slo uses the REST API directly and has no MCP tools, while some skills can generate JSON offline.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/kubesense-ai/kubesense-mcp-skills --skill kubesense-skills

After installing, type/kubesense-skillsin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

README.md
package-lock.json
drafts/
README.md
kubesense-investigate.md
kubesense-alerts/references/
datadog-migration.md
import-json.md
kubesense-dashboards/references/
variables-and-rows.md
panel-config.md
kubesense-mcp/references/
multi-query.md
field-catalog.md
kubesense-metrics/references/
cloud-metric-catalog.md
metric-catalog.md
kubesense-slo/references/
api-payloads.md
datadog-migration.md
kubesense-sql/references/
recipes.md

Details

From the project README

KubeSense Skills

Observability skills for KubeSense, grouped by telemetry surface.

Skills

SkillDescription
kubesense-mcpThe tool layer — connection, auth, the full tool inventory, the WHERE/field-name contract every query skill inherits. Start here.
kubesense-logsSearch and aggregate logs
kubesense-tracesSpans, latency percentiles, error rates, distributed-trace waterfalls
kubesense-metricsPromQL/MetricsQL over Kubernetes, infrastructure and cloud-provider (AWS/GCP/Azure/Atlas/Confluent/Kong) metrics
kubesense-splSPL — the piped log query language, and its inverted field contract
kubesense-sqlRaw ClickHouse SQL over logs and traces — joins, CTEs, window functions
kubesense-infraCluster inventory: clusters, nodes, pods, workloads, infra failures, recent deploys
kubesense-alertsList, investigate, and create alert rules; generate import JSON; migrate Datadog monitors
kubesense-dashboardsCreate dashboards, or generate preset JSON to import
kubesense-sloCreate and manage service level objectives; error budgets, burn-rate alerting, Datadog SLO migration

Routing

The user is asking…Skill
"show me errors / grep the logs"kubesense-logs
"p99 latency / slow requests / which service is failing"kubesense-traces
"CPU, memory, disk, saturation, capacity"kubesense-metrics
"how is my RDS / EC2 / Azure VM / Atlas cluster doing"kubesense-metrics
"what's running / why is this pod restarting / what changed"kubesense-infra
"alert me when… / what's firing / convert this Datadog monitor"kubesense-alerts
"build me a dashboard"kubesense-dashboards
"track an SLO / error budget / port a Datadog SLO"kubesense-slo
a query needing a join, CTE, or window functionkubesense-sql
"search the logs with a pipeline" / SPL tabkubesense-spl
"this alert fired — what's wrong?"kubesense-alerts, then kubesense-infra
a tool returned a field-name or WHERE errorkubesense-mcp

Prerequisites

Every skill except kubesense-dashboards, kubesense-alerts and kubesense-slo (which can generate JSON offline) needs the KubeSense MCP server connected. Those three need it for their validate and create tools, and for discovering real metric and field names — offline they fall back to JSON built from names the user supplies. kubesense-slo has no MCP tools at all: it drives the REST API directly, and uses the MCP query tools only to discover names.

The MCP server is served by kubeapi at /mcp over Streamable HTTP and authenticates with the same credentials as the REST API.

claude mcp add --scope user --transport http kubesense \
  https://<your-kubesense-host>/mcp \
  --header "x-api-key: <your-api-key>"

See kubesense-mcp for auth alternatives and verification.

Two Rules That Apply Everywhere

  1. Discover before querying. Never invent a field name, metric name, cluster name, or label. Every surface has a discovery tool; a guessed name either errors or — worse — returns an empty result that reads like "zero".

  2. Field names are catalog labels, not storage columns. Logs/traces queries take type, instance, service, method, status_code — not level, pod_name, app_service, subtype, return_code. The alert engine accepts a different set. See kubesense-mcp for the contract and kubesense-alerts for the divergence.

Install

npx skills add kubesense-ai/kubesense-mcp-skills --full-depth -y

Important

--full-depth is required. Without it the installer stops at this root SKILL.md and installs only this index — 1 skill instead of 11 — with no warning.

If you are reading this as the only installed KubeSense skill, that is what happened. Re-run the command above with --full-depth.

Verify with npx skills add kubesense-ai/kubesense-mcp-skills --list --full-depth, which should report 11 skills. Or pick individual skills — see the README.

Comments

Markdown supported