Cloud Native MCP Server

A production-grade MCP server for Kubernetes and cloud-native infrastructure management, exposing 15 services and 327 tools across SSE and Streamable HTTP modes.

GitHub Repository Quick Start

Core Value#

Single Operations Interface

Unifies Kubernetes, Helm, Grafana, Prometheus, Kibana, and more behind one MCP surface, reducing context switching for operators and agents.

Production Security Controls

Supports apikey / bearer / basic authentication, rate limiting, and audit logging for enterprise hardening and compliance workflows.

Agent-Friendly Output

Pagination and summarization patterns help AI assistants stay efficient during large incident investigations.

Typical Use Cases#

Incident Triage

Correlate pod state, events, logs, and metrics to shorten the path from alert to root cause.

Release and Change Control

Use Helm and Kubernetes tools for rollout, rollback, scaling, and controlled production changes with auditable traces.

Cross-Observability Analysis

Bridge Prometheus, Grafana, Loki, Jaeger, Sentry, and OpenTelemetry signals for end-to-end diagnostics.

Project Snapshot#

Integrated Services

327

MCP Tools

Run Modes

MIT

License

Integrated Services#

Kubernetes 34 tools

Core orchestration and resource management workflows.

Helm 34 tools

Package lifecycle and release operations for Kubernetes apps.

Argo CD 7 tools

Application, project, cluster, and manifest inspection for GitOps workflows.

Grafana 55 tools

Dashboards, alerting, and visualization management.

Prometheus 20 tools

Metrics querying, rules inspection, and monitoring workflows.

Loki 7 tools

LogQL queries, label discovery, and stream inspection for log-first troubleshooting.

Kibana 73 tools

Log exploration and analytics for Elastic-based observability.

Elasticsearch 12 tools

Index inspection, search, and cluster operation support.

Alertmanager 16 tools

Alert routing, silence management, and incident visibility.

Jaeger 8 tools

Distributed tracing and request-path diagnostics.

Nacos 9 tools

Namespaces, configs, service discovery, instances, and node-level inspection.

Langfuse 37 tools

LLM observability for traces, sessions, prompts, scores, metrics, projects, memberships, and keys.

Sentry 9 tools

Error monitoring for issue triage, issue events, and project discovery.

OpenTelemetry 12 tools

Telemetry pipeline checks for traces, logs, and metrics.

Utilities 6 tools

General-purpose helpers for day-to-day operational tasks.

Quick Start#

Docker

1
2
3
4
5
6
7
8
docker run -d \
  --name cloud-native-mcp-server \
  -p 8080:8080 \
  -v ~/.kube:/root/.kube:ro \
  -e MCP_AUTH_ENABLED=true \
  -e MCP_AUTH_MODE=apikey \
  -e MCP_AUTH_API_KEY='ChangeMe-Strong-Key-123!' \
  mahmutabi/cloud-native-mcp-server:latest

Binary

1
2
3
curl -LO https://github.com/mahmut-Abi/cloud-native-mcp-server/releases/latest/download/cloud-native-mcp-server-linux-amd64
chmod +x cloud-native-mcp-server-linux-amd64
./cloud-native-mcp-server-linux-amd64 --mode=sse --addr=0.0.0.0:8080

Source

1
2
3
4
git clone https://github.com/mahmut-Abi/cloud-native-mcp-server.git
cd cloud-native-mcp-server
make build
./cloud-native-mcp-server --mode=streamable-http --addr=0.0.0.0:8080

Availability Check#

1
2
3
4
5
# 1) Health check
curl -sS http://127.0.0.1:8080/health

# 2) End-to-end SSE handshake + initialize check (run at repo root)
make sse-smoke BASE_URL=http://127.0.0.1:8080

Common Entry Points#

Aggregate SSE endpoint (--mode=sse): http://127.0.0.1:8080/api/aggregate/sse
Aggregate Streamable HTTP endpoint (--mode=streamable-http): http://127.0.0.1:8080/api/aggregate/streamable-http
Health endpoint: http://127.0.0.1:8080/health

Pre-Production Checklist#

Authentication and Access

Enable `MCP_AUTH_ENABLED=true` in production.
Choose one mode: `apikey`, `bearer`, or `basic`.
Apply least-privilege access to Kubernetes and external systems.

Observability and Audit

Enable structured logs and core metrics collection.
Enable audit logs if change tracking is required.
Continuously validate `/health` and core upstream service checks.

Performance and Resilience

Tune rate limits, timeouts, and concurrency for your traffic profile.
Prefer summary and pagination tools to limit context size.
Load-test with realistic multi-service tool-call patterns.

Documentation Map#

FAQ and Troubleshooting Entry#

Getting Started FAQ

Answers common implementation questions around auth mode, transport mode, client integration, and production rollout strategy.

Read FAQ

Troubleshooting Playbook

Step-by-step checks for startup failures, 401 responses, SSE handshake issues, and unavailable service integrations.

Open Troubleshooting