UniFi

@UniFi list all devices at our main office site and show me which ones haven't checked in during the past 24 hours

Found 47 devices across your main office site. 44 devices are online and reporting normally. 3 devices have not checked in within the last 24 hours: (1) AP-Conference-Room-B (U6-Pro, last seen 26 hours ago), (2) SW-Warehouse-02 (USW-24-POE, last seen 31 hours ago), (3) AP-Loading-Dock (UAP-AC-LR, last seen 38 hours ago). All three devices are showing as 'disconnected' in the controller. The warehouse switch may need physical inspection given its extended offline period.

This example combines List Devices with timestamp filtering to surface connectivity issues. Requires API key with read access to the target site. Device state reflects the most recent controller sync — if the UniFi controller itself is behind on polling, results may lag real-world status by several minutes.

@UniFi get the current SD-WAN configuration status for our multi-site setup and tell me if there are any tunnel failures or warnings

Retrieved status for SD-WAN config 'Multi-Site-Primary' (ID: sdwan_abc123). Deployment generation 47 is active across 5 sites. Tunnel health: 18 of 20 tunnels are up. 2 tunnels show degraded status: Site A ↔ Site D (packet loss 4.2%, latency spikes to 180ms), Site B ↔ Site E (intermittent drops, 3 reconnects in past hour). No critical errors. One warning flagged: Site C gateway firmware is one version behind the others, which may cause policy sync delays.

This example demonstrates Get SD-WAN Config Status for proactive network monitoring. Useful for diagnosing multi-site connectivity before users report issues. Note that tunnel metrics reflect the controller's last measurement cycle (typically 1-5 minutes). Write access is not required, but you need visibility into the SD-WAN config itself.

@UniFi pull ISP performance metrics for the last 7 days across all our sites and identify which location has the worst uplink reliability

Analyzed ISP metrics across 5 sites over the past 7 days. Site rankings by uplink reliability: (1) HQ - 99.97% uptime, avg latency 12ms, (2) Branch-North - 99.89% uptime, avg latency 18ms, (3) Branch-West - 99.76% uptime, avg latency 22ms, (4) Branch-South - 98.34% uptime, avg latency 31ms, (5) Branch-East - 96.12% uptime, avg latency 47ms. Branch-East experienced 11 outage events totaling 6.8 hours downtime, with 4 events exceeding 30 minutes. Peak packet loss during business hours averaged 2.1%. Consider escalating with the ISP or evaluating a secondary uplink for this location.

This example showcases Get ISP Metrics with historical aggregation, letting the AI synthesize reliability trends from raw telemetry. The MCP returns time-series data; the AI's summary makes it actionable. Metrics depend on the UniFi gateway's WAN monitoring being enabled at each site. Large date ranges may hit API rate limits on accounts with many sites.

When this MCP wins for distributed IT teams

A 6-person IT team managing UniFi gear across 12 retail locations hits a Friday afternoon outage. The on-call engineer pulls ISP metrics and device status across all sites in one Switchy thread instead of logging into each controller separately. This MCP is the right call when you're already on UniFi and your team needs to correlate network events across multiple sites without switching contexts. The API key setup is straightforward if you're a super admin. The threshold: if you're troubleshooting a single site or your team doesn't have owner-level access, the web console is faster. For multi-site correlation during incidents, this MCP cuts your mean-time-to-context from 15 minutes to under two.

When this MCP handles post-deployment checks

A 3-person MSP team deploys SD-WAN configs for a client with four branch offices and needs to verify tunnel status and error states before signing off. They use the SD-WAN config and status tools in a Switchy workspace to pull deployment health across all sites in one pass, then document findings in the same thread. This MCP is ideal when you're managing SD-WAN at small-to-medium scale and need programmatic access to topology and tunnel metrics. The trade-off: if you're only checking one or two configs occasionally, the UniFi console is less setup overhead. For teams running repeatable validation workflows across multiple client networks, this MCP turns a 20-minute checklist into a 3-minute query.

When this MCP automates monthly network reports

A 4-person managed services provider sends monthly uptime and performance reports to 8 clients, each running UniFi at 2-5 locations. They use this MCP to pull ISP metrics and device lists by host, then generate summaries in Switchy without exporting CSVs or screenshotting dashboards. This MCP is the right fit when you're already using UniFi for client networks and need structured data for reporting or SLA tracking. The boundary: if your clients are on mixed vendors or you need sub-hourly granularity, you'll hit the API's metric limits. For monthly or weekly reporting at MSP scale, this MCP replaces manual data collection and keeps the entire workflow in one shared workspace.