About talos-mcp-server
talos-mcp-server is an MCP server for operating Talos Linux clusters through typed actions backed by the official machine API. It covers common cluster setup and health tasks while retaining a raw talosctl option for less common operations.
Key features
- 01Generate cluster configuration
- 02Apply machine configuration to nodes
- 03Bootstrap etcd and retrieve kubeconfig
- 04Validate cluster health
- 05Run raw talosctl arguments
Use cases
- 01Create Talos cluster configuration
- 02Apply configuration and bootstrap a control plane
- 03Retrieve kubeconfig for a cluster
- 04Check cluster health or run less common operations
MCP support
Exposes a single `talos` tool with typed actions to MCP clients. It uses `talosctl`, the official gRPC client for the Talos machine API.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMETalos MCP Server
A Model Context Protocol server for Talos Linux cluster operations — one thin tool, typed actions, zero runtime dependencies.
Companion to virtualbox-mcp-server: that server provisions the nodes, this one turns them into a Kubernetes cluster.
Design
One tool (talos) with typed actions for the cluster hot path — no CLI syntax needed:
| Action | Does |
|---|---|
gen_config | Generate cluster config + talosconfig (cluster, endpoint, optional output_dir) |
apply_config | Apply machine config to a node (node, file, insecure for maintenance mode) |
bootstrap | Bootstrap etcd on the control plane (once per cluster) |
kubeconfig | Retrieve the cluster kubeconfig |
health | Cluster health validation |
config_merge | Merge a talosconfig context (never overwrites ~/.talos/config) |
version | Client or node version |
run | Raw talosctl argv for the long tail (etcd, logs, reboot, upgrade…) |
resolve | Report platform/install state without downloading or running |
Backend: talosctl — the official gRPC client of the Talos machine API. An existing PATH install is always preferred; otherwise the correct binary for the host's OS/architecture is auto-installed from the official Sidero release, verified against GitHub's asset digest.
The full cluster recipe ships to clients in the MCP initialize handshake, so a fresh session can bring up a cluster without external knowledge.
Prerequisites
None beyond a hypervisor for the nodes. The server is a single Go binary; talosctl is auto-installed if absent.
Installing
Published on the official MCP Registry
as io.github.bryanjbelanger/talos-mcp-server — that name is what
registry-aware clients install by
(current entry).
For hosts that support MCP Bundles, download the .mcpb from the
latest release
and open it. One bundle covers every platform; it carries a binary per
(os, arch) and picks one at startup. The server needs no configuration —
guest credentials and endpoints are per-call parameters, and talosctl is
auto-installed on first use if the host has none.
Otherwise take the bare binary for your platform from the same release, or build it:
go build -o talos-mcp-server .
Registration with Claude Code
claude mcp add --scope user talos -- /path/to/talos-mcp-server
Verify with claude mcp list (✔ Connected), then restart the session.
Bundled skill
skills/talos-cluster — the end-to-end procedure: provision nodes via the virtualbox MCP server, configure and bootstrap via this one. Install by copying (or symlinking) into ~/.claude/skills/.
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.