NetApp/mcp

Give AI agents useful, governed access to your data estate
NetApp Model Context Protocol (MCP) servers connect AI assistants and agents to
technical documentation, storage operations, infrastructure telemetry, and
enterprise data. Use this catalog to choose the server that matches the outcome
you need—not simply the product you own.
New to MCP servers?
MCP is an open standard that gives AI applications a consistent way to connect
to external data, tools, and workflows. Start with the vendor-neutral
official MCP introduction
for the fundamentals. For optional hands-on practice, see DeepLearning.AI's
*MCP: Build Rich-Context AI Apps with Anthropic*;
it is supplemental training, not NetApp product documentation.
Choose by outcome
| What you want an AI agent to do | Start with | Why |
| --- | --- | --- |
| Administer ONTAP across one or more clusters | ONTAP MCP Server | Broad ONTAP administration across NAS, SAN, NVMe, data protection, and core storage services |
| Identify infrastructure problems, investigate performance, and plan capacity | Harvest MCP Server | Uses current and historical Harvest telemetry from Prometheus or VictoriaMetrics across ONTAP, E-Series, and StorageGRID; Cisco Nexus adds ecosystem visibility |
| Manage Google Cloud NetApp Volumes end to end | Google Cloud NetApp Volumes MCP Server | Broad GCNV control-plane coverage, including pools, protection, security, NAS, and iSCSI |
| Provision, clone, and protect datasets or workspaces | NetApp DataOps Toolkit MCP server family | Four task-focused MCP servers for ONTAP, Azure NetApp Files, Google Cloud NetApp Volumes, and Kubernetes, with self-service workflows for developers, data scientists, and platform teams |
| Ground answers and workflows in official NetApp technical documentation | docs.netapp.com MCP server | Focused remote service for searching and retrieving published documentation and generating RAG-based answers without a custom integration |
| Review risks, security recommendations, upgrade readiness, and AutoSupport for your NetApp systems | Active IQ MCP Server | Uses the AutoSupport data your systems already send to NetApp, so there is nothing to deploy or integrate; covers risks, security recommendations and remediation, support cases, contracts, and upgrade guidance |
| Investigate infrastructure health, performance, capacity, alerts, logs, and collection status in Data Infrastructure Insights | DII MCP Server | Brings tenant-scoped DII inventory, telemetry, events, alerts, collector health, and notification posture into natural-language, read-only investigations |
| Turn NetApp AIQ Digital Advisor data into specialist health, troubleshooting, and remediation guidance | Digital Coach Model Context Protocol (MCP) Server | Synthesizes Digital Advisor, hardware, knowledge, documentation, EMS, and AutoSupport context for expert-level reasoning instead of returning only raw records |
MCP server details
ONTAP MCP Server
**Move from storage advice to storage action.** Give AI agents a validated,
centrally managed path to discover and operate ONTAP instead of relying on
generated scripts or making users translate recommendations into manual steps.
- **Use it when:** Storage and infrastructure teams need broad administration,
multi-cluster workflows, or a shared MCP service for ONTAP.
- **Coverage:** API discovery and read access, volumes, SVMs, qtrees, snapshots,
snapshot policies, SnapMirror, QoS, NFS, SMB/CIFS, LUNs and igroups, iSCSI,
FCP, NVMe, DNS, and network interfaces. A read-only mode is available when
agents should inspect but not change the environment.
- **Runs as:** A self-hosted Streamable HTTP service, distributed as a container
image and Helm chart.
- **Works with:** NetApp AFX, AFF, FAS, Amazon FSx for NetApp ONTAP, Cloud
Volumes ONTAP, and ONTAP Select.
- **Start here:** Documentation ·
Source
Harvest MCP Server
**Turn infrastructure telemetry into decisions.** Ask which systems need
attention, where latency is rising, or when capacity will run out—without
manually navigating dashboards or writing PromQL. Harvest spans ONTAP, E-Series,
and StorageGRID, with Cisco Nexus providing additional ecosystem visibility, so
each answer draws on portfolio-wide telemetry history collected by Harvest and
stored in the deployment's Prometheus or VictoriaMetrics database. That observed