Cloudback MCP Server

About Cloudback MCP Server

Cloudback MCP Server is an official add-on for the Cloudback backup platform. It connects AI coding assistants and chat clients-Claude Code, Claude Desktop, Cursor, VS Code, or any stdio MCP client-to Cloudback's configuration layer. Through plain-language prompts, users can inspect, filter, and bulk-edit backup definitions, schedules, storages, and retention policies across GitHub, GitLab, Azure DevOps, and Linear accounts.

Review

Cloudback MCP Server turns backup configuration into a chat-driven workflow. Instead of clicking through a dashboard or writing one-off scripts, operators can ask questions about backup coverage and push bulk changes across hundreds of repositories in a single prompt. The server intentionally stops short of touching backup data itself-restores, on-demand runs, and deletions stay in the dashboard.

Key Features

• List and filter backup definitions across multiple accounts and platforms from a single MCP client, with case-insensitive matching.
• Bulk-update schedules, storage targets, and retention policies by name. The assistant reads existing retention policy names and reuses them, avoiding hard-coded day counts.
• Read and adjust account-level defaults: timezone, default schedule, default storage, and default retention.
• Separate read-only inspection tools from write tools. Every write call is scoped to one account (one API key), requires explicit definition IDs, and is surfaced for client approval before execution.
• Runs as a Docker image (myrtlelabs/cloudback-mcp) with a three-step setup: create an Operations API key per account, populate a single appsettings.json, and register the server in the MCP config.

Pricing and Value

The MCP server itself is free. It works with an existing Cloudback account; every Cloudback plan starts with a free trial. Specific plan pricing beyond the trial is not detailed in the launch materials. The server does not carry a separate cost-it's included as an interface for Cloudback's backup service.

Pros

• Bulk configuration changes happen in one prompt after listing the exact set of definitions, which reduces the chance of scope mistakes.
• Account-scoped API keys and definition-ID-level writes keep bulk edits contained; no wildcard operations are allowed.
• Write calls require client-side approval, with the account, definition IDs, and new values shown before anything runs.
• All changes land in Cloudback's audit log, creating a record of what was applied and when.
• Inspection-focused design: you must list before you edit, which encourages verification before mutation.

Cons

• The write response does not bundle before/after values for each changed definition. To confirm the exact delta, you need to re-read definitions after the write.
• No single-step dry-run or diff receipt across all definitions. Today, the "before" view comes from the List call, and the "after" requires a separate read.
• Not suited for teams that want to trigger restores, delete backup data, or run on-demand backups through chat-those operations are deliberately excluded from the MCP surface.

This tool fits platform engineers and DevOps teams already using Cloudback who want a faster path to bulk config changes and natural-language inspection. It's less relevant for organizations that need chat-driven restore or data deletion, since those actions remain dashboard-only. For audit-heavy environments, the lack of a bundled diff in the write response means operators will lean on re-reads and the dashboard audit log to confirm changes.