Skip to main content
Glama
vbr-write-mcp

vbr-write-mcp

by Aquatick

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
MCP_TOKENSNoJSON object mapping long random bearer tokens to team names, used for HTTP transport authentication.
VBRW_DRY_RUNNoSet to true to avoid sending any write requests and only show the requests that would be sent.
VBR_BASE_URLYesThe URL of the Veeam Backup & Replication REST API server, for example http://host:9419.
VBR_PASSWORDYesThe VBR password used to authenticate with the Veeam Backup & Replication REST API.
VBR_USERNAMEYesThe VBR username used to authenticate with the Veeam Backup & Replication REST API.
VBRW_AUDIT_LOGNoPath for the JSONL audit log. Every attempt is recorded with plan, confirmation type, result, error, and client details, with secrets masked.
VBRW_HTTP_PORTNoHTTP port used when VBRW_TRANSPORT=http, for example 8001.
VBRW_TRANSPORTNoTransport mode. Set to http for shared HTTP usage; otherwise the default stdio transport is used.
VBRW_TOOL_GROUPSNoNarrows the exposed tool groups.
VBRW_ACCESS_LEVELNoAccess level for the server. Use write to hide destructive tools and reject dangerous variants of mixed tools; recommended for production.
VBRW_BLOCKED_PATHSNoAdditional paths blocked for api_write. The paths /api/oauth2/* and registerVbr are always blocked.
VBRW_DISABLED_TOOLSNoComma-separated or otherwise configured list of tool names to disable.
VBRW_CONFIRMATION_MODENoServer-side confirmation mode. Valid values are auto (default), elicit, token, and none. The none mode disables confirmation and is not recommended.auto

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

Write and administration tools for one Veeam Backup & Replication (VBR) server over its REST API (1.3-rev2). They complement Veeam's read-oriented vbr-mcp: use vbr-mcp (or api_read here) to find ids and check state, and these tools to change things.

Every state-changing tool goes through a server-side safety layer:

  • Destructive tools (delete, restore, failover/failback, stopping sessions, edits of existing objects, settings, removing licenses, marking malware clean...) are confirmed by the server. Either the user is asked directly by the MCP client, or the tool returns status: confirmation_required with a plan and a confirm_token. In that case show the plan to the user in plain words (target name, what will happen, the impact lines), ask for an explicit yes, and only then call the same tool again with the same arguments plus confirm_token. Never invent or reuse a token, never treat the user's original request as the yes, and never confirm on their behalf.

  • status: dry_run means the server is in dry-run mode and nothing was sent.

  • status: submitted means VBR accepted the request. Most actions are asynchronous: say they were started, then follow the returned session with get_session_result.

Creating or configuring something for the user (a backup job, repository, credentials, ...): never ask the user for JSON or field names. Ask in plain language for what is missing, a few questions at a time, and offer concrete choices read from the server instead of open questions (the VMs/vApps to protect from the inventory, the repositories with their free space, the proxies, existing credentials and encryption passwords). For a backup job that is: name, what to protect, target repository, retention, schedule, application-aware processing and guest credentials, encryption, and whether a backup copy or GFS is wanted; propose sensible defaults the user can accept with one word. Then build the body from describe_endpoint / describe_model, show a short plain-language summary of what will be created (not JSON), and call the tool only after the user agrees.

Before any action, resolve ids with read tools; never guess or construct an id. Edits (update_object, update_settings) take only the fields to change and merge them into the current object; lists are replaced as a whole, so send the complete new list. When you need a request body you do not know, call describe_endpoint (and describe_model for nested or polymorphic parts, see x-subtypes) instead of guessing field names. api_write is the escape hatch for any write endpoint without a dedicated tool. When VBR refuses a call, report its reason and do not repeat the same call unchanged. A question about an action ("what happens if...") is not a request to run it.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": true
}
logging
{}
prompts
{
  "listChanged": true
}
resources
{
  "subscribe": false,
  "listChanged": true
}

Tools

Functions exposed to the LLM to take actions

NameDescription
set_protection_group_enabledB

Enable or disable a protection group. A disabled group stops discovery and agent deployment for its computers (disabling is confirmed by the server).

rescan_protection_groupB

Rescan a protection group now: discover new computers and refresh agent state. Returns a session.

manage_discovered_computersA

Act on computers a protection group discovered (ids from vbr-mcp get_protection_group_entities).

Additive actions run directly: install/upgrade agent, install CBT driver, rescan, trust, prepare for remote bare-metal recovery, recreate embedded recovery media. Confirmed by the server: uninstall agent / CBT driver / all components, reboot, remove (removes one computer from the group - pass exactly one id).

set_agent_backup_attachmentB

Attach existing backups to a managed agent computer (e.g. after reinstalling it), or detach them.

reboot_recovery_appliancesB

Reboot Veeam Agent recovery appliances (one by id, or all when appliance_ids is omitted).

delete_backupA

Delete a whole backup (all restore points of every machine in it) from disk.

include_gfs=True also deletes GFS (weekly/monthly/yearly) restore points - without it they are kept. from_configuration_only=True removes the backup only from the VBR configuration (fromDB) and leaves the files on the repository (they can be imported again). Immutable restore points cannot be deleted before their immutability expires; VBR then reports an error. Irreversible.

delete_backup_objectA

Delete one machine's restore points from a backup ("Delete from disk" for a single VM/computer).

object_id is the backup object id from vbr-mcp get_backup_details(include objects). Irreversible.

move_backupA

Move a backup (whole chain) to another repository; the job is re-pointed to the target.

Runs as a move session and can take long; jobs using the backup cannot run meanwhile. If it stops halfway, manage_move_copy_session decides between retry, forget and undo.

copy_backupsA

Copy backups to another repository (or, for machine backups, a folder) - the source stays as it is.

spec is the request body: for unstructured_data a CopyBackupSpec (see describe_model); for machines VBR does not publish the schema - typically backupIds and a target repository or folder. Check describe_endpoint('/api/v1/backups/machines/copyTo', 'POST').

move_backup_objects_to_jobB

Re-home machines' backups from one job's backup to another job (e.g. after moving VMs between jobs), so the target job continues their chains. Without backup_object_ids all objects of the backup move.

run_backup_health_checkB

Run a storage-level health check (CRC/consistency) on backups, optionally repairing what it finds.

Give backup_ids and/or job_ids. repair=True rewrites corrupted blocks / marks broken points and is confirmed as destructive; repair_options is a HealthCheckRepairOptions object. Returns a session.

set_backup_file_malware_stateA

Mark one backup file (restore point file) as Clean or Infected for malware detection.

Marking infected blocks normal restores from it (secure restore only); marking clean lifts that, so it is confirmed by the server.

manage_move_copy_sessionA

Resolve a backup move/copy session that is waiting for a decision: Retry the failed part, ForgetFailed (keep what moved, drop the rest) or StopAndUndo (roll back).

create_recovery_mediaC

Create Veeam Recovery Media (bare-metal recovery ISO) for an agent backup object.

describe_endpointA

Look up the VBR REST API 1.3-rev2 reference bundled with this server.

With a path and method: the endpoint's summary, description, path/query parameters and request-body schema (nested models inlined to depth; deeper ones appear as {"$ref": Model} - expand them with describe_model). A model with x-subtypes is polymorphic: its type field selects one of the listed subtypes, whose extra fields the body must carry. With search words: matching endpoints (method, path, summary).

describe_modelB

Schema of one VBR API model by name (e.g. EntireCloudDirectorVMCustomizedRestoreSpec, VmwareObjectModel).

Shows fields, which are required, enum values and descriptions; x-subtypes lists the concrete variants of a polymorphic model.

api_readA

Send a GET request to the VBR REST API (read-only), e.g. to find ids or check state after an action.

path has its ids filled in (e.g. /api/v1/jobs/{id}); params holds query parameters (filters, skip/limit, orderColumn/orderAsc). If Veeam's vbr-mcp is connected too, its dedicated list tools give richer answers.

get_session_resultA

State and result of a VBR session - the asynchronous run started by almost every action tool.

With wait_seconds, polls every 5 s until the session reaches Stopped or the time is up. When the session stopped with a Warning or Failed result (or include_logs and it is still running), the session log records are included so the reason is visible.

api_writeA

Send any state-changing request to the VBR REST API - for write endpoints no dedicated tool covers.

Examples: High Availability cluster switchover, Azure / file-share / FCD Instant Recovery, Entra ID restores, disk publishing (Data Integration API), helper appliances, KMS servers. Look the endpoint up with describe_endpoint first and build body from its schema. Only endpoints of the VBR 1.3-rev2 API are accepted. Always treated as destructive: the server asks for confirmation before sending. reason is recorded in the audit log.

rescan_managed_serversB

Rescan several managed servers at once, or all of them when managed_server_ids is omitted. Returns a session.

rescan_inventoryA

Refresh the cached virtualization inventory (VMs, vApps, datastores, networks) of one managed server, e.g. after a VM or vApp was created and does not show up in browse_inventory yet.

set_default_mount_serverC

Make a mount server the default one for its repositories' restores (FLR, Instant Recovery).

get_server_certificateA

Fetch a server's TLS certificate thumbprint or SSH fingerprint - needed in the spec when adding a managed server with create_object(kind='managed_server'). Read-only (changes nothing in VBR).

discover_cloud_director_vcentersA

List the vCenter Servers attached to a VMware Cloud Director server, before adding it (CloudDirectorHostDiscoverySpec: see describe_model). Read-only.

clone_jobA

Clone a job. VBR creates the copy under a derived name; rename or adjust it afterwards with update_object(kind='job', ...) and check its schedule before it first runs.

Supported job types: VSphereBackup, HyperVBackup, CloudDirectorBackup, BackupCopy, WindowsAgentBackup, LinuxAgentBackup. Returns the new job's model.

start_quick_backupA

Quick Backup: an on-demand incremental of one machine that is already protected by a backup job.

object is the inventory object exactly as VBR returns it: for vsphere/cloud_director a VmwareObjectModel (platform, hostName, name, type: VirtualMachine, objectId) from vbr-mcp browse_inventory; for hyperv a HyperVObjectModel; for agent an AgentObjectModel (id, name, type, protectionGroupId, platform: Agent). The machine must have a backup in a job; VBR uses that job's settings and adds a restore point to its chain. Returns a session.

apply_job_configurationB

Push a Veeam Agent backup policy's current configuration to the agents it manages ("Apply configuration").

Only for agent backup policies (*AgentBackupServerPolicy / *WorkstationPolicy job types).

clear_job_cacheA

Clear the backup cache of a Veeam Agent backup policy (the local cache agents write to while the repository is unreachable). Make sure cached restore points were uploaded first.

install_licenseB

Install a license file on the backup server, replacing the installed license. Confirmed by the server.

refresh_licenseA

renew: download the renewed license from Veeam for the installed one; update: fetch the current license state (e.g. after a contract change) from the Veeam licensing server. Needs internet access on the server.

set_license_optionsA

Switch license options: automatic license update (with proactive support) and instance consumption by unlicensed agents. Only the options you pass are changed.

manage_instance_licenseB

Assign an instance license to a workload as Server or Workstation, or remove the workload's instance license (it then needs a license again at the next backup).

remove_installed_licenseB

Remove the installed license (or only its Instance or Socket section). Jobs stop processing unlicensed workloads. Confirmed by the server.

mark_backup_objects_cleanA

Mark backup objects (machines) as clean after a malware investigation - typically a false positive.

mark_restore_points_clean also clears the Suspicious/Infected flags of their restore points, so normal restores are allowed again. exclude_from_detection stops future detection for them (a security exception - use with care). reason is mandatory here and is stored by VBR and in the audit log.

manage_malware_exclusionsA

Create/update (upsert, with exclusions - a list of MalwareDetectionCommonSpec, see describe_model) or delete (with exclusion_ids) malware-detection exclusions for backup objects. Exclusions weaken detection, so both are confirmed by the server.

start_encryption_analysisB

Run an encryption analysis for a malware event (checks whether data in the backup was encrypted, e.g. by ransomware). Event ids: vbr-mcp list_malware_events.

report_malware_eventB

Create a malware event from an external source (EDR/XDR/SIEM): VBR then flags the machine's restore points.

machine is a SuspiciousActivityMachineSpec (see describe_model; identifies the machine by FQDN/IP/UUID).

set_ioc_monitoringA

Enable or disable monitoring of specific Indicators of Compromise (IoC) in file-level malware detection.

indicators names the IoC entries to switch (as listed by api_read('/api/v1/malwareDetection/settings/fileDetection/indicatorsOfCompromise') or the VBR console). Disabling weakens detection and is confirmed by the server.

create_objectA

Create a VBR object from a full spec (the POST body of its collection).

Kinds: job, managed_server, repository, scale_out_repository, proxy, mount_server, credentials, cloud_credentials, encryption_password, kms_server, protection_group, recovery_token, unstructured_data_server, entra_id_tenant, ad_domain, global_vm_exclusion, role, user, datastore_latency. Look the body up first with describe_endpoint('<collection path>', 'POST'); most specs are polymorphic on type (see x-subtypes, e.g. VSphereBackupJobSpec/BackupJobSpec for job, WindowsLocalStorageSpec/LinuxHardenedStorageSpec/SmbStorageSpec for repository, ViHostSpec/CloudDirectorHostSpec for managed_server, StandardCredentialsSpec/LinuxCredentialsSpec for credentials). Additive, so no confirmation is needed by default; adding servers and repositories runs as a session.

update_objectA

Edit an existing VBR object: pass only the fields to change in changes; the tool reads the object, deep-merges changes into it and sends the complete model (VBR PUT replaces the whole object). Nested objects merge field by field; lists are replaced as a whole, so send the full new list. replace=True sends changes as the complete body instead.

Kinds: job, managed_server, repository, scale_out_repository, proxy, mount_server, credentials, cloud_credentials, encryption_password, kms_server, protection_group, recovery_token, unstructured_data_server, entra_id_tenant, ad_domain, global_vm_exclusion, role, user, datastore_latency (not ad_domain). user edits the role list, global_vm_exclusion the note, encryption_password the hint. The confirmation plan shows a field-level diff. Example: disable job encryption -> kind='job', changes={'storage': {'advancedSettings': {'storageData': {'encryption': {'isEnabled': false}}}}}. Edits can change what is protected and how: confirmed by the server.

delete_objectA

Remove a VBR object.

Kinds: job, managed_server, repository, scale_out_repository, proxy, mount_server, credentials, cloud_credentials, encryption_password, kms_server, protection_group, recovery_token, unstructured_data_server, entra_id_tenant, ad_domain, global_vm_exclusion, role, user, datastore_latency (not mount_server). Optional options are the DELETE query parameters: deleteBackups (repository, scale_out_repository - also deletes the backups stored there), uninstallEverything (protection_group - uninstalls agents). Use delete_backup for backups. Confirmed by the server.

update_settingsA

Edit server-wide VBR settings (or a per-object settings block) with patch semantics: pass only the fields to change; the tool reads the current settings, merges and sends the full model, and the confirmation plan shows the diff. Lists are replaced as a whole.

Areas: config_backup, general_options, email, notifications, event_forwarding, host_authentication, storage_latency, node_exporter, vsa_event_forwarding, vsa_event_forwarding_daemon, traffic_rules, malware_general, malware_suspicious_files, security_mfa, security_analyzer_schedule, repository_access_permissions, hyperv_server_volumes, azure_instant_recovery_switchover, file_share_instant_recovery_switchover. Areas whose path has an {id} need id (repository, Hyper-V server or Instant Recovery mount). Look the model up with describe_endpoint('<path>', 'PUT') when unsure of field names. Example: config backup retention -> area='config_backup', changes={'restorePointsToKeep': 30}.

list_replicasA

List VM replicas, or - with replica_id - that replica's restore points (newest first), to pick the replicaPointIds for failover/failback. Read-only.

failover_replicasA

VMware vSphere replica failover. Pass one replica point (restore point of a replica) per VM.

  • now: power the replicas on at the DR site (disaster: production is lost).

  • planned: last sync, shut production down, switch to the replicas (migration / expected outage).

  • permanent: make the replicas the permanent production VMs (no failback).

  • undo: discard the failover - replicas are powered off, their changes lost. After now/planned, finish with failback_replicas or failover_replicas(mode='permanent'|'undo').

failback_replicasA

Failback of failed-over vSphere replicas, in steps:

  1. start - sync replica changes back to target (OriginalVM: the original VM, OriginalLocation: re-create it where it was, Customized: elsewhere). switchover decides when production takes over (Auto, Manual then switch_to_production, or Scheduled; change it later with change_switchover_time).

  2. commit - finalize; or undo - return to the failed-over state.

rescan_replicasB

Rescan replicas to refresh their state in VBR (all replicas when replica_ids is omitted).

restore_cloud_director_vmA

Full restore of a VMware Cloud Director VM from a restore point of a Cloud Director backup job.

OriginalLocation restores the VM into its original vApp, replacing the existing VM if it is still there. Customized restores it into vapp under new_name (required, keeps the original untouched), optionally with network mapping and datastore/storage policy. Inventory objects (vApp, networks, datastores) come from vbr-mcp browse_inventory on the Cloud Director server. proxy_ids pins backup proxies (default: automatic). Returns a restore session.

restore_hyperv_vmA

Full restore of a Microsoft Hyper-V VM. OriginalLocation overwrites the VM in place; Customized restores to destination_host (and optionally a new name, storage and network mapping - see describe_model for HvRestoreTargetDatastoresSpec / HvRestoreTargetNetworkSpec / HvRestoreTargetNameSpec).

start_instant_recovery_hyperv_vmA

Instant Recovery to Microsoft Hyper-V: run the VM straight from the backup, without a full data copy.

OriginalLocation publishes it in place of the original; Customized on destination_host; OtherPlatform recovers a VMware/agent/other backup onto Hyper-V (needs destination_host). Finish with finish_instant_recovery_hyperv_vm (migrate to production or unmount).

finish_instant_recovery_hyperv_vmA

Finish a Hyper-V Instant Recovery mount: migrate the running VM to production storage, or unmount (tear it down; changes made inside the recovered VM are lost). Mount ids: api_read('/api/v1/restore/ instantRecovery/hyperV/vm').

start_file_level_restoreA

Mount a machine restore point for guest file-level restore (FLR). Returns the mount session; its id is the session_id for browse_file_level_restore, restore_files and unmount_file_level_restore.

Mounting changes nothing in production. Linux FLR uses a helper appliance/host configured in VBR.

browse_file_level_restoreC

List a folder (or search) in a file-level-restore mount. Read-only.

restore_filesA

Restore files/folders from a file-level-restore mount.

original: back to the same machine and path. restore_type: Keep keeps existing files and writes the restored ones next to them under a RESTORED- name, Overwrite replaces existing files, PermissionOnly restores only permissions. another: to target_host / target_path. credentials_id: guest credentials if VBR cannot use the job's.

unmount_file_level_restoreB

Unmount a file-level-restore mount when you are done with it.

change_credentials_secretA

Rotate the secret of a stored credentials record used by VBR to reach servers, guests or cloud accounts.

password (standard or Linux account), linux_root_password, linux_private_key (secret = the key, optional passphrase), cloud_secret_key (secret key of a cloud credentials record). Every job and server using the record uses the new secret from the next connection - a wrong value breaks them.

change_encryption_passwordA

Change an encryption password (and its hint). Jobs using it encrypt new backups with the new password; existing backups still need the old one to restore - keep it safe.

verify_encryption_passwordA

Check whether password matches a stored encryption password (e.g. before a restore). Changes nothing.

manage_user_securityB

Reset a user's MFA (they must enroll again at next login), or switch the service-account mode (service accounts are exempt from MFA - use for automation accounts only). Roles: update_object(kind='user').

clone_roleB

Clone a custom role; adjust the copy with update_object(kind='role', ...).

security_analyzerB

Security & Compliance Analyzer: start a scan; suppress one best-practice finding (with a note - it stops being reported, so it is confirmed); reset one finding's status or reset_all. Schedule: update_settings(area='security_analyzer_schedule').

stop_sessionA

Stop a running session of any type: a restore, Instant Recovery, rescan, health check, SureBackup, malware scan, move/copy, etc. (for a backup job run, vbr-mcp stop_job gives graceful-stop control).

What is already done is not rolled back; a stopped restore can leave a partially restored VM or files behind.

stop_taskC

Stop one task (/api/v1/tasks/{id}), e.g. a long-running export, deployment or recovery-media task.

stop_automation_sessionB

Stop an automation (import/export) session started by import_configuration or export_configuration.

export_configurationA

Export VBR configuration objects as JSON import specs - for documentation, review, migration to another VBR server, or as a template for create_object. Exporting credentials / cloud credentials / encryption passwords is confirmed by the server because the output is sensitive.

import_configurationA

Import configuration objects (as produced by export_configuration, e.g. on another VBR server). Objects that exist already may be updated. Returns an automation session; stop it with stop_automation_session. Confirmed by the server.

export_support_logsB

Collect a support log bundle (for a Veeam support case) for some jobs or hosts. With date_from/date_to (ISO-8601) only that range is exported, otherwise everything. Returns a log-export session.

send_test_emailB

Send a test message with the configured email notification settings (update_settings(area='email')).

high_availability_clusterA

Backup server High Availability cluster: planned switchover to the secondary node (ignore_lag skips the database replication lag check), or failover when the primary node is lost. The REST API and this MCP server reconnect to the new primary only if VBR_BASE_URL points at the cluster's shared name.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

TDQS

B3.2/5.0

Scored across 68 tools

Disambiguation4/5

Most tools target clearly distinct resource+action pairs (delete_backup vs delete_backup_object vs delete_object, move_backup vs copy_backups vs move_backup_objects_to_job, start/finish Instant Recovery pairs). A few boundaries blur: stop_session vs stop_task vs stop_automation_session, the generic create_object/update_object/delete_object family vs the dedicated create_* tools, and api_write vs any dedicated write tool. Descriptions largely resolve these by explicitly stating scope (api_write is 'for write endpoints no dedicated tool covers').

Naming Consistency3/5

All names are snake_case, but verb conventions vary widely (set_, manage_, rescan_, start_, apply_, mark_, clone_, verify_) and a few tools are noun-only or noun-first (api_read, api_write, security_analyzer, high_availability_cluster, describe_endpoint). Patterns are still readable and mostly verb_noun, but not a single predictable scheme across 68 tools.

Tool Count2/5

68 tools is far beyond a well-scoped set, driven by exposing granular operations for a very large VBR domain. While many tools are individually justified, the count creates a heavy selection burden, and generic escape hatches (api_write, create_object/update_object) coexist with numerous dedicated variants, which is exactly the strain that pushes this dimension down.

Completeness4/5

Coverage is broad: jobs, repositories, protection groups, backup deletion/move/copy, multiple restore types, malware/IoC handling, licensing, settings, security, sessions, HA and config import/export. Minor gaps remain, notably no dedicated vSphere VM full restore (only Cloud Director and Hyper-V restores) and no dedicated job-run tool (only quick backup), but api_write and companion vbr-mcp tools provide workarounds.

Maintenance

ActivityMaintained
ResponsivenessNo recent community issues