Changelog

Custom dashboards just got a lot more flexible. Every dashboard and every individual widget can now be scoped to a specific set of monitors, so a single dashboard can show a focused view for one chain, one region, or one team. This release also brings quick actions to the monitor list, a mobile-friendly redesign across monitor screens, and a fix for gas price widgets misreporting during RPC outages.

• FeatureDashboard-level monitor filtering: scope an entire dashboard to a chosen set of monitors
• FeaturePer-widget monitor filtering: override the dashboard scope for any individual widget
• FeatureDuplicate and delete quick actions added directly to the monitor list
• ImprovementLatency sparklines now render on mobile monitor cards
• ImprovementReworked mobile layout for monitor creation, monitor detail, and dashboard editing screens
• FixFixed gas price widgets showing a false 0.0 Gwei reading when an RPC provider was temporarily unreachable
• FixFixed incident notification links pointing to a broken route
• FixFixed invite email uniqueness check incorrectly applying across organizations instead of within one

The BlackTide AI SRE Agent is now generally available. When a monitor fires, the agent automatically investigates the incident: it queries on-chain data, correlates with recent deployments, checks gas price spikes, and surfaces a probable root cause before your on-call engineer opens their laptop. Average time-to-diagnosis drops from 18 minutes to under 90 seconds.

• FeatureAI SRE Agent with automatic root cause analysis on every P1 and P2 incident
• FeatureOn-chain context fetcher: correlates incidents with recent block activity, mempool congestion, and gas price anomalies
• FeatureDeployment correlation: links incidents to commits pushed within the configurable lookback window (default 2h)
• FeatureNatural language incident summary delivered to Slack thread and PagerDuty incident notes
• ImprovementIncident timeline UI redesigned to surface AI findings prominently alongside manual observations
• ImprovementAgent runs in read-only mode by default - no automated remediation without explicit opt-in

BlackTide now supports Cardano-based chains and Cosmos-based modular chains as first-class monitor targets. Both chain families support block height lag detection, validator status checks, and custom alert thresholds. Additionally, all webhook deliveries are now HMAC-SHA256 signed by default to prevent spoofing.

• FeatureCardano-base monitoring: block height lag, node health, Substrate JSON-RPC checks
• FeatureCosmos-base monitoring: block height, data availability sampling health, validator set changes
• FeatureHMAC-SHA256 webhook signing via X-BlackTide-Signature header - enabled by default for all new webhook channels
• FeatureWebhook delivery log: per-event history of all delivery attempts with response codes and latency
• ImprovementCosmos SDK monitor executor refactored to support new chain onboarding without code changes
• ImprovementChain selector search now fuzzy-matches on chain ID and EVM hex chain ID
• FixFixed block height comparison edge case when Cosmos-base DA blocks have variable finalization times

Managing on-call schedules just got dramatically easier. This release ships a full on-call rotation builder directly inside BlackTide, bidirectional Opsgenie integration, and a configurable escalation policy engine that can chain Slack, PagerDuty, Opsgenie, and webhook actions with time-based escalation steps.

• FeatureOn-call rotation builder: define weekly schedules with team member assignments and automatic handoff notifications
• FeatureOpsgenie integration: bidirectional incident sync with full escalation policy support
• FeatureEscalation policy engine: chain up to 5 escalation steps with configurable delays (5 min, 15 min, 30 min, 1h)
• FeatureOverride management: add one-off schedule overrides without modifying the base rotation
• ImprovementAlert channel configuration page redesigned with drag-and-drop step ordering
• ImprovementPagerDuty integration now supports severity mapping: P1 → high urgency, P2/P3 → low urgency
• FixFixed race condition where rapid monitor state flaps could create duplicate PagerDuty incidents
• FixResolved timezone handling bug in on-call schedule display for UTC+12 and UTC-12 team members

A focused security release hardening the BlackTide platform against web application attacks. This ships CSRF double-submit cookie protection on all state-changing API endpoints, a strict Content Security Policy, and a full audit log of all team member actions accessible to workspace admins.

• FeatureCSRF double-submit cookie protection on all POST/PATCH/DELETE endpoints
• FeatureContent Security Policy header with strict source allowlists and nonce-based script execution
• FeatureAudit log: timestamped record of all monitor create/edit/delete, team member changes, and alert rule modifications
• FeatureSession management UI: view and revoke active sessions per device
• ImprovementhttpOnly auth cookies now have SameSite=Strict for additional CSRF protection
• ImprovementAdded HSTS header with 1-year max-age and includeSubDomains
• FixResolved reflected XSS vector in monitor name display on public status pages