Overview
longhorn-mcp is an MCP server for the Longhorn distributed storage manager API. It offers full read/write coverage of Longhorn's REST API, spanning volumes, snapshots, backups, backing images, nodes, settings, engine images, recurring jobs, and more. It supports a --read-only mode, both stdio and Streamable HTTP transports, and is distributed with an official Helm chart published alongside every release.
The server can be started locally with npx or deployed in-cluster via the Helm chart. In --http mode it serves Streamable HTTP on a configurable host and port, with an --allowed-hosts allowlist for Host/Origin checking. Read-only mode is enforced in two ways: write tools are never advertised in tools/list, and the client layer refuses any non-GET request regardless of which tool triggered it.
Every Longhorn resource returned by the server embeds an actions map of the RPC actions currently valid for it. The server rewrites each action URL into the MCP tool that performs it, along with that tool's parameter shape, so a client can call it without a separate schema lookup. Tool descriptions generally end with a link to the relevant page in Longhorn's official documentation, pinned to the Longhorn version the default target runs.
Destructive or hard-to-reverse tools carry an advisory instruction not to be called unless the user has explicitly authorized that specific action in the current conversation. This guidance is model-facing and not enforced; the --read-only flag is described as an enforced, structural guarantee.
Pricing
Free
Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.
Visit WebsitePrices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.
Use Cases
- Teams use longhorn-mcp to give AI assistants read and write access to Longhorn volumes, snapshots, backups, and other storage resources.
- Operators use the --read-only mode to expose Longhorn to an LLM session without allowing any write operations.
- Teams deploy the server in-cluster using the official Helm chart.
- Users connect MCP clients such as Claude Code, Claude Desktop, Cursor, and VS Code to a local stdio server.
- Teams front --http mode with an MCP-aware gateway for authentication, rate limiting, and routing.
- Developers run scripts/introspect-schema.ts to cross-check the package's tool coverage against a live Longhorn manager.
Features
Exposes Longhorn API through MCP
Manages volumes and snapshots
Handles backups and backing images
Accesses nodes and engine images
Manages settings and recurring jobs
Restricts operations with read only mode
Supports stdio and Streamable HTTP
Provides an official Helm chart
Reviews
No reviews yet. Be the first to share your experience!