BTRFS Snapper MCP

December 30, 2025 ยท View on GitHub

PyPI Built with Claude Code

An MCP (Model Context Protocol) server for managing BTRFS snapshots via Snapper and monitoring BTRFS filesystem health. Exposes two unified tools with action-based dispatch to minimize tool proliferation while providing full snapshot management and disk health monitoring.

Features

  • Two-tool design: snapper for snapshots, btrfs_health for filesystem monitoring
  • Full Snapper support: list, create, delete, rollback, cleanup, diff, status
  • BTRFS health monitoring: usage, device stats, scrub status, balance status
  • Flexible configuration: Environment variables for cross-system portability
  • No hardcoded paths: Works with any Snapper configuration and mount points

Installation

pip install snapper-mcp

Using uvx (no install required)

uvx snapper-mcp

From source

git clone https://github.com/danielrosehill/BTRFS-Snapper-MCP.git
cd BTRFS-Snapper-MCP
pip install -e .

Configuration

The server is configured via environment variables, making it portable across systems:

VariableDefaultDescription
SNAPPER_MCP_USE_SUDOtrueSet to false if user has direct permissions
SNAPPER_MCP_SNAPPER_PATHsnapperPath to snapper binary
SNAPPER_MCP_BTRFS_PATHbtrfsPath to btrfs binary
SNAPPER_MCP_SUDO_PATHsudoPath to sudo binary
SNAPPER_MCP_DEFAULT_CONFIGrootDefault snapper config name
SNAPPER_MCP_DEFAULT_MOUNT/Default BTRFS mount point
SNAPPER_MCP_TIMEOUT60Command timeout in seconds

MCP Client Configuration

Claude Code / Claude Desktop

Add to your MCP settings (e.g., ~/.claude/settings.json or Claude Desktop config):

{
  "mcpServers": {
    "btrfs-snapper": {
      "command": "uvx",
      "args": ["snapper-mcp"]
    }
  }
}

With custom configuration

{
  "mcpServers": {
    "btrfs-snapper": {
      "command": "uvx",
      "args": ["snapper-mcp"],
      "env": {
        "SNAPPER_MCP_DEFAULT_CONFIG": "root",
        "SNAPPER_MCP_DEFAULT_MOUNT": "/",
        "SNAPPER_MCP_USE_SUDO": "true"
      }
    }
  }
}

Tools

snapper - Snapshot Management

Manages BTRFS snapshots via Snapper.

Actions

ActionDescriptionRequired Parameters
configsList available snapper configurations-
listShow all snapshots for a configconfig (optional)
createCreate a new snapshotconfig, description (optional)
deleteRemove a snapshot by numberconfig, snapshot_id
rollbackRestore to a previous snapshotconfig, snapshot_id
cleanupPrune snapshots using an algorithmconfig, cleanup_algorithm
diffShow file differences between snapshotsconfig, snapshot_id, snapshot_id_end
statusList changed files between snapshotsconfig, snapshot_id, snapshot_id_end

Examples

// List configurations
{"action": "configs"}

// List snapshots
{"action": "list", "config": "root"}

// Create snapshot
{"action": "create", "config": "root", "description": "Before system update"}

// Delete snapshot
{"action": "delete", "config": "root", "snapshot_id": 5}

// Rollback (requires reboot)
{"action": "rollback", "config": "root", "snapshot_id": 3}

// Cleanup with algorithm
{"action": "cleanup", "config": "root", "cleanup_algorithm": "number"}

// Compare snapshots
{"action": "diff", "config": "root", "snapshot_id": 1, "snapshot_id_end": 5}

Cleanup algorithms:

  • number: Keep a fixed number of snapshots
  • timeline: Keep snapshots based on age (hourly, daily, weekly, monthly, yearly)
  • empty-pre-post: Remove empty pre/post snapshot pairs

btrfs_health - Filesystem Health Monitoring

Monitors BTRFS filesystem health and status.

Actions

ActionDescriptionParameters
usageComprehensive filesystem space usagemount_point (optional)
devicesList devices in the BTRFS arraymount_point (optional)
statsDevice error statisticsmount_point, device (optional)
scrub_statusStatus of scrub operationmount_point (optional)
balance_statusStatus of balance operationmount_point (optional)
dfSpace allocation by data typemount_point (optional)
filesystem_showFilesystem informationmount_point (optional)

Examples

// Check filesystem usage
{"action": "usage", "mount_point": "/"}

// Check for device errors
{"action": "stats", "mount_point": "/"}

// Check scrub status
{"action": "scrub_status", "mount_point": "/"}

// Show data/metadata allocation
{"action": "df", "mount_point": "/"}

// Show filesystem info
{"action": "filesystem_show"}

Prerequisites

  • Linux with BTRFS filesystem
  • Snapper installed and configured (for snapshot management)
  • Python 3.10+
  • Sudo access (unless running as root or with appropriate permissions)

Installing Snapper

Ubuntu/Debian:

sudo apt install snapper

Fedora/openSUSE:

sudo dnf install snapper
# or
sudo zypper install snapper

Creating Snapper Configs

# For root filesystem
sudo snapper -c root create-config /

# For home partition
sudo snapper -c home create-config /home

Sudo Configuration

For passwordless operation, add to /etc/sudoers.d/btrfs-snapper:

yourusername ALL=(ALL) NOPASSWD: /usr/bin/snapper
yourusername ALL=(ALL) NOPASSWD: /usr/bin/btrfs

Or disable sudo if you have direct permissions:

{
  "env": {
    "SNAPPER_MCP_USE_SUDO": "false"
  }
}

Adapting for Other Systems

Different Snapshot Tools

The architecture can be adapted for other snapshot tools (e.g., Timeshift, ZFS snapshots) by:

  1. Replacing the run_snapper_command() function with your tool's CLI
  2. Adjusting the action handlers for your tool's command syntax
  3. Updating environment variables as needed

Different Filesystems

For ZFS or other filesystems:

  1. Replace run_btrfs_command() with your filesystem's CLI
  2. Update BtrfsAction enum with relevant actions
  3. Implement new handlers for your filesystem's commands

The action-dispatch pattern remains the same regardless of underlying tools.

License

MIT