Robbie Guardian transforms ServerDesk from a reactive tool into a proactive infrastructure intelligence platform. Guardian monitors CPU, memory, disk, containers, deployments, logs, and more — detecting meaningful trends, generating predictive insights, triggering automatic investigations, and delivering a full notification inbox with a daily morning briefing.
Changes in this release
- Guardian snapshot engine: lightweight infrastructure snapshots per connection, up to 100 retained per server
- Trend analysis: linear regression across 8 metrics — CPU, memory, disk, error logs, docker instability, diagnostics, git drift, deployment failures
- Predictive insights: disk-full ETA, memory exhaustion timeline, container instability forecast, branch drift prediction
- 12 deterministic Guardian rules: high CPU sustained, memory critical, disk critical, container restart loops, deployment-error cascade, error log spike, git drift, WordPress security updates, connection health error, database risk signals, memory leak trend, disk filling fast
- Automatic investigation trigger: rules fire Robbie investigation (rate-limited to once per 30 minutes) — results saved to history and inbox
- Robbie Inbox: notification centre with categories (investigation, warning, recommendation, prediction, briefing), priority levels, mark read/pin/dismiss, unread count badge
- Guardian Feed: full timeline of infrastructure events grouped by date with severity-coloured dots
- Morning Briefing: daily summary card — health score, highlights, investigation count, warning count
- Health trend cards: per-metric trend direction, magnitude, confidence score, delta
- 8 new analytics events for the Guardian lifecycle
- Help Centre topic covering Guardian monitoring, inbox, trends, predictions, and rules
- All monitoring is deterministic, read-only, and consumes zero AI credits
*Released 5 July 2026 · Category: AI*