The dashboard your team uses today and the agent operations of tomorrow run on the same Domotz platform — same Layer-2 inventory, same RBAC, same $1.50 per managed device. No premium AI tier. No markup on AI service consumption.
14-day free trial. No credit card required. Under 15 minutes to deploy.
Get a balanced summary from your AI client of choice. Prompt seeded with the facts that matter — agent-native, vendor-built, OAuth-scoped, $1.50 per device.
The Domotz MCP Server is an agent-native, vendor-built MCP server that lets AI agents and IT teams operate the same networks from the same platform. It exposes Domotz monitoring, configuration, and remediation capabilities through the open Model Context Protocol standard, governed by OAuth-scoped consent and the same RBAC compliance that governs the Domotz web app. It is included at $1.50 per managed device per month — the standard Domotz platform price — with no premium AI tier and no markup on AI service consumption.
Domotz works on both sides of the timeline. The Domotz web application gives technicians the visibility, configuration, and remediation they need today, with accurate Layer-2 inventory and full multi-vendor coverage. The Domotz MCP Server gives AI agents the same operational capability through the same platform, so when a team is ready to investigate, configure, or remediate through an agent, the platform is already there.
Rich Domotz MCP tooling spanning inventory, metrics, topology, configuration backup, alert rules, credentials, sensors, scripts, and remote actions transforms tedious manual processes into single-action workflows. Twenty clicks become one prompt — across every site you monitor, not just the flagship ones.
IT professionals prompt their managed networks in natural language and receive direct answers and actions through the Domotz MCP Server. Investigate device status, surface alerts, and pull metrics without learning where every view and filter lives in the dashboard.
AI agents speak in operator language — devices, profiles, sensors, alerts, credentials — to investigate networks, query metrics, view topology, and compare configuration backups across customers and sites. Twenty clicks become one prompt, or a scheduled autonomous routine.
AI agents apply device profiles, attach sensors and scripts, set credentials, save alert rules, manage license state (managed vs unmanaged devices), and execute remote actions such as cycling outlet power, restarting devices, running scripts, and resolving alerts — with full guardrails.
Read access is available immediately. Any agent action that changes state on the network requires explicit account-level opt-in by an administrator before it can be invoked.
Each action is OAuth-scoped to a specific capability, and sensitive operations require client confirmation at the moment of execution. Tokens are short-lived and per-client.
The MCP Server enforces the same RBAC and user-management policies as the Domotz web app. Agents operate strictly within the permissions of the authenticating user.
Every action through the Domotz MCP Server is captured in the Domotz platform audit trail, including failed and denied attempts. The full record is reviewable from the web app, alongside technician-initiated actions.
Administrators turn AI automation on with the same governance they already use today.
Built and supported
Domotz designs, ships, and supports the MCP server as part of the platform — security-reviewed, version-controlled, and backed by the same Domotz support relationship that backs the web app.
Replaces
Other approaches leave teams calling no one when something breaks. The Domotz MCP Server is shipped and supported as part of the platform — designed, security-reviewed, and version-controlled by Domotz.
Real operational capability
The Domotz MCP Server exposes operational capability across investigation, reporting, configuration, and remediation — purpose-built so AI agents and IT operators work in natural language alongside the existing developer API.
Replaces
Read-only or surface-level integrations that can answer questions but cannot complete the work.
Pricing
The MCP Server is part of the standard Domotz platform price. No new SKU, no per-seat add-on, no markup on AI service consumption. You control the AI spend.
Replaces
A premium AI tier, a per-seat AI add-on, or a vendor markup on AI consumption that gates experimentation behind a procurement cycle.
Five capability groups. Every row visible. Function names are real MCP tools your AI client invokes when you prompt.
device_inventory
device_performance
get_uptime
Real-time visibility into device status and availability — faster issue detection and reduced downtime.
get_device_interfaces
get_interface_traffic
Track bandwidth usage, errors, and interface health to identify bottlenecks and performance issues.
collector_overview
collector_discovery_status
Monitor network environments at the site level for centralized multi-site visibility.
attach_sensor
list_device_sensors
list_preconfigured_sensors
Collect custom metrics — CPU, memory, temperature — for deep device insight beyond basic monitoring.
list_device_profiles
apply_device_profile
Standardize monitoring across device types for scalable, repeatable monitoring setup.
network_topology
Discovered device-to-device links and upstream dependency map to improve resilience and uptime awareness.
get_device_metrics
list_device_metrics
list_metric_functions
Time-series metrics with summary or chronos-bucket views, per-device metric catalogs, and the evaluation functions every alert rule binds to.
search_devices
list_device_types
device_inventory
update_device_metadata
Full visibility into all network assets for accurate asset tracking and management.
get_account_info
search_organizations
search_collectors
Manage multiple customers and sites for MSP scalability and operational control.
list_tags
create_tag
Organize and segment infrastructure for faster filtering and troubleshooting.
get_device_credentials
set_device_credential
Manage credentials across SNMP, SSH, HTTP/HTTPS, TELNET, WINRM, and ONVIF. Returns STATUS only — passwords, private keys, and community strings never leave the platform.
restart_device
cycle_outlet_power
Restart devices; cycle a single outlet on a PoE switch or PDU for power-cycling when other recovery paths are exhausted.
list_scripts
list_device_scripts
attach_script
execute_script
Attach and run custom drivers. Script execution is destructive on the target device and requires explicit confirmation.
update_device_monitoring_state
Set a device to MANAGED or UNMANAGED. Hits the same $1.50/managed device licensing quota the dashboard uses.
get_config_backup_status
list_config_backup_history
Track backup success and history to ensure recoverability and reliability.
get_config_backup_entry
Access historical device configurations for troubleshooting and rollback support.
compare_config_backup
Identify changes between configurations to detect unauthorized or risky changes.
compare_config_backup
list_config_backup_history
Detect unauthorized or unexpected changes to strengthen security posture.
search_devices
device_inventory
list_device_types
Identify all devices on the network to reduce blind spots and unknown assets.
get_collector_credential_coverage
Collector-wide credential audit across SNMP, OS monitoring, configuration management, ONVIF, and device management — no blind spots, no secrets returned.
device_inventory
network_topology
Power source and upstream dependency map for resilience planning.
search_alerts
resolve_alert
get_device_alerts
Route alerts into external systems for faster incident response.
create_alert_rule
list_alert_rules
list_metric_functions
list_communication_channels
Discover existing rules, create new ones with required functions and channels — channel_ids are returned only at creation, so reuse over create.
attach_alert_rule
bind_alert_rule_to_device
bind_alert_rule_to_agent
get_device_alert_rule_bindings
get_agent_alert_rule_bindings
list_variable_alert_rule_candidates
Idempotent binding — ALREADY_BOUND returns rather than errors. Dedupe before creating.
The MCP server is included in the standard Domotz platform price — no premium AI tier, no per-seat add-on, and no markup on AI service consumption. The dashboard your team uses today and the agent operations of tomorrow run on the same platform, at the same $1.50 per managed device per month.
See pricingThe Domotz MCP Server is an agent-native, vendor-built MCP server that lets AI agents and IT teams operate the same networks from the same platform. It exposes Domotz monitoring, configuration, and remediation capabilities through the open Model Context Protocol standard, governed by OAuth-scoped consent and the same RBAC compliance that governs the Domotz web app.
The Domotz MCP Server is included in the standard $1.50 per managed device per month price. There is no separate AI tier, no per-seat add-on, and no markup on AI service consumption. The dashboard your team uses today and the agent operations of tomorrow run on the same platform at the same price.
The Domotz MCP Server works with any AI client that supports the Model Context Protocol, including ChatGPT, Claude, Perplexity, and Gemini. Built on an open standard, it is LLM-vendor agnostic, so teams can keep using the AI client they already standardize on.
Every action through the Domotz MCP Server is governed by the same RBAC and user-management policies that govern the web app. Write processes require account opt-in. Each action is OAuth-scoped, with client confirmation for sensitive operations. Administrators turn AI automation on with the same governance they already use today.
AI agents can query inventory, investigate metrics and topology, compare configuration backups, apply device profiles, attach sensors and scripts, set credentials, save alert rules, manage license state (managed vs unmanaged devices), and execute remote actions such as cycling outlet power, restarting devices, running scripts, and resolving alerts. Every action runs inside the trust model described above.
The Domotz MCP Server is vendor-built and vendor-supported — designed, shipped, and maintained as part of the Domotz platform. It exposes real operational capability across monitoring, configuration, and remediation through the open MCP standard, so teams can use it from the AI clients they already trust without giving up the governance the platform already provides.
Yes. The same prompt that investigates one network can investigate every network. Agent-native operations scale across every site you monitor, with no extra per-site cost. Smaller sites get the same agent operability as flagship accounts, billed at the same $1.50 per device.
Yes. The Domotz platform audit trail captures every attempt — not only successful actions — for accountability and forensic review. The audit trail is accessible through the Domotz web app, exactly as it is for technician-initiated actions.
No. The Domotz MCP Server runs on the same platform, the same data, and the same pricing as the Domotz web app. There is no separate AI product to buy, no migration project, and no retrofit. Both the technician dashboard and agent operations share the same Layer-2 inventory.
Start a free trial, connect a collector, and grant the MCP server access via OAuth in your client of choice. Whether you operate through ChatGPT, Claude, Perplexity, or Gemini, the same controls govern what the agent can read or change. Full setup instructions live in the Domotz help center: Domotz MCP Server setup guide.