Unitrends MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Unitrends MCP Servershow me open backup alerts and this week's success rate"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Unitrends MCP Server
An MCP server for the REST API of a Unitrends backup appliance (Recovery Series / Unitrends Backup). It lets an AI agent read the systems an appliance manages, protected assets, running and finished jobs, restorable backups, open alerts and backup success rates and, after explicit confirmation, queue a restore.
Transport: MCP Streamable HTTP at
POST /mcp, health probe atGET /healthStateless and multi-tenant: credentials arrive on every request as HTTP headers and are never stored; the appliance login token is obtained per request and discarded with it
Python 3.12, uv, FastMCP, httpx
License: Apache-2.0, Copyright 2026 MSPbots.ai
MSPbots Connector
The MCP Gateway built for MSPs: visible, controlled, and your data stays where it belongs.
Get started at connect.mspbots.ai →
MSPbots Connector gives your AI clients one MCP URL to reach your entire MSP stack (RMM, PSA, documentation, security, finance), and gives you control over who can call which tool and a record of everything that was called.
Hosted service: connect.mspbots.ai · Free to use.
Why Connector

Visible | Every tool call is logged: who, which vendor, when. Usage dashboard, CSV export, SIEM streaming. |
Controlled | Role-based, per-tool allowlist. Deny by default. |
Protected | Credentials encrypted per record. AI clients never see your vendor passwords or tokens. |
MSP-native | Built for the MSP stack only: RMM, PSA, documentation, security, finance. |
Quick start
Sign up at connect.mspbots.ai.
Connect your vendors. Pick from 150+ connectors and authorize each one on the vendor's own page, or bring your own API keys. No extra vendor accounts needed.
Set access. Create roles and decide which tools each role may call. Nothing is allowed until you allow it.
Add your MCP URL to your AI client. Copy the URL from your Connector dashboard into Claude, ChatGPT, Cursor, VS Code, Gemini, Codex, or any MCP-capable client.
Ask your AI to do real work, and watch every call appear in the audit log.
Features
Unified gateway
One MCP URL for all your AI clients
150+ connectors across the MSP stack
Custom MCP servers supported, free
Authorization happens on each vendor's own page; the AI never sees passwords or tokens
Bring your own credentials: use the vendor API keys you already have
Multiple accounts per app; connect once, reuse across AI clients
Tool search: finds the right action for the task
One sign-in; vendor credentials are injected behind the scenes and OAuth tokens refresh automatically
Visibility
Audit log for every tool call (who, which vendor, when)
Usage dashboard: call volume, top vendors, error rate, per-user activity
Export audit logs to CSV, or stream to Loki, Graylog, or LogScale (SIEM)
Optional recording of tool arguments and result summaries (off by default, enabled per organization)
Control
Per-role, per-tool allowlist, deny by default
owner / admin / member roles
SCIM provisioning with Entra ID, Okta, JumpCloud, and Google Workspace
Security
Credentials encrypted one record at a time (AES-256-GCM, per-record key)
AI clients authenticate with OAuth 2.1 + PKCE
Vendor responses are scanned for prompt-injection patterns and returned as structured data. This is one layer of defense, not a guarantee against every attack.
Static egress IP for customer firewall and API allowlisting
Pricing
Free.
Links: connect.mspbots.ai · mspbots.ai
Related MCP server: mcp-cohesity
Authentication headers
Header | Required | Type | Default | Allowed values | Description | Example |
| Yes | string | none | https URL | Root URL of the appliance (a bare host name gets |
|
| Yes | string | none | - | Appliance user name |
|
| Yes | string | none | - | Password of that user |
|
| No | string |
|
| Set |
|
Where to get them:
Use (or create) a dedicated user in the appliance's user settings (menu names vary by appliance version). A role that can view systems, assets, jobs, backups and alerts covers the read tools; restores need a role that may run restores.
The base URL is the address you open the appliance's web UI with. The appliance must be reachable over HTTPS from the environment that runs this server.
Appliances ship with a self-signed certificate; either install a trusted certificate or send
X-Unitrends-Verify-TLS: false.
How credentials are sent to Unitrends: the server calls POST {base}/api/login with the user
name and password, then sends every API call with the returned token in the AuthToken
header. A 401 triggers one fresh login. Values that look like unresolved templates (${...})
are treated as missing.
A request to /mcp without the three required headers gets HTTP 401 naming the missing
ones. A non-https URL or an invalid X-Unitrends-Verify-TLS value does not block the request
(so tools/list works with any values); every tool call then returns an invalid_argument
error naming the header, and nothing is sent to the appliance. An invalid
X-Unitrends-Verify-TLS value never turns TLS verification off.
Tools
All tools return a JSON string. applianceId is a system id from
unitrends_list_appliances; omitting it where it is optional covers every system the login
can see. Dates are ISO 8601 (2026-06-01 or 2026-06-01T00:00:00Z).
Tool | Purpose | Parameters (required in bold) |
| List the systems (appliances) this login manages: id, name, role, version, status, capacity | - |
| One system's status, role, capacity and hardware details | applianceId |
| Protected assets (clients: servers, VMs, hosts) with OS, agent version and type | applianceId |
| One protected asset: OS, agent, applications | applianceId, assetId |
| Jobs running or queued now, with percent complete | applianceId, type |
| Finished jobs, newest first (default last 7 days) | since, until, applianceId, type, limit |
| An asset's restorable backups: backup id, type, date, status, size, storage | assetId, applianceId, since, until |
| Restore a backup onto a system (write, needs | recoveryPointId, targetAssetId, targetPath, applianceId, confirm |
| A restore job: running (percent complete) or finished (result) | restoreId, applianceId |
| Open alerts, newest first | applianceId, severity ( |
| Backup success rate for a period: totals and per-asset counts, failing assets first | since, until, applianceId |
type for jobs is one of backup, restore, replication, archive, ir, flr,
certification, backup-copy.
Errors come back as a JSON envelope instead of an exception:
{"error": {"code": "unauthorized", "message": "Unitrends rejected the username or password (login failed)", "retryable": false}}code is one of not_configured, unauthorized, not_found, invalid_argument,
rate_limited, upstream_error; messages from the appliance's result array are passed
through. Reads are retried with backoff on 429, 5xx and network errors; the restore request
is never retried.
Run locally
uv sync
uv run python -m unitrends_mcp # listens on 0.0.0.0:8080or with Docker:
docker build --platform linux/amd64 -t unitrends-mcp:dev .
docker run --rm -p 8080:8080 unitrends-mcp:devConfiguration (environment variables, all optional): MCP_HTTP_HOST (default 0.0.0.0),
MCP_HTTP_PORT (default 8080), LOG_LEVEL (default INFO). Credentials are never read
from the environment.
Testing example
# list tools
curl -s http://localhost:8080/mcp \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-H "X-Unitrends-Base-URL: https://backup.example.com" \
-H "X-Unitrends-Username: your-username" \
-H "X-Unitrends-Password: your-password" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}'
# backup success rate for the last 7 days
curl -s http://localhost:8080/mcp \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-H "X-Unitrends-Base-URL: https://backup.example.com" \
-H "X-Unitrends-Username: your-username" \
-H "X-Unitrends-Password: your-password" \
-H "X-Unitrends-Verify-TLS: false" \
-d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"unitrends_get_success_rate","arguments":{}}}'Unit tests (no network access needed):
uv run pytestDelivery notes
Built on the documented appliance API. Every call uses the endpoints in the official Unitrends API documentation:
/api/loginwith theAuthTokenheader,/systems,/clients,/jobs/active,/jobs/history,/backups/catalog,/restore/full,/alerts/openand/summary/status.Tool count (11). Tool names and required parameters are unchanged; some tools gained optional filters (
applianceId,type,limit,severity).Restore confirmation. Older versions asked for confirmation with an interactive prompt, which a stateless gateway cannot relay, so restores could not run there. Now
unitrends_queue_restorereturns a preview unless called withconfirm=true; the tool description tells the agent to set it only after the user approves that exact backup and target. It is the only write tool and is annotated destructive.Result size exception (SOP 2.5).
unitrends_list_job_historyandunitrends_list_alertshave no paging in the API, so they keep a maximum of 1000 (default 50); results are still cut to 20,000 characters with a truncation notice.Date ranges default to the last 7 days instead of prompting; job history honours the requested range.
License
This server cannot be deployed
Maintenance
Related MCP Connectors
Manage, check, investigate and automate server, database and storage backups from your AI assistant.
- RowsafeOAuthsh.rowsafe
Database backups and an AI agent safety net: restore points, health, audited changes as you.
Read-only search of 2,189 AI agent tools (MCP servers, x402 APIs) with daily automated trust scores.
Unified MCP Server is a remote MCP connector for AI agents and vertical AI products that provides access to 22,000+ authorized SaaS tools across 400+ integrations and 24 categories directly inside LLMs (Claude, GPT, Gemini, Cohere). Tools operate only on explicitly authorized customer connections, enabling agents to safely read and write against live third-party systems.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables AI agents to securely access and manage Commvault environments, including job control, backup schedules, client/storage management, SLA monitoring, and optional DocuSign envelope backup integration.17Apache 2.0
- FlicenseAqualityCmaintenanceProvides AI assistants with access to Cohesity DataProtect for managing backups, recovery operations, and data protection monitoring via the Cohesity REST API. It enables users to trigger on-demand backups, monitor cluster health, and manage protection policies through natural language commands.149 npm2-
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to interact with the Netdisco network management platform through its complete REST API, supporting device inspection, port queries, VLAN searches, and job management.MIT
- AlicenseNot gradedqualityAmaintenanceEnables AI assistants and automation agents to inspect Proxmox clusters, nodes, storage, containers, VMs, and recent events through the Proxmox VE API, with optional backup actions.5Apache 2.0