Release v0.7.0

InsightFlare v0.7.0 focuses on improving interface design, expanding analytics capabilities, and increasing the efficiency of the Dashboard and edge runtime at scale. This release redesigns API v1, adds advanced filtering with nested combinations of complex conditions, saved filters, traffic channel analytics, improved Realtime data transport, and a versioned traffic blocking-rule system.

Across the Dashboard, v0.7.0 expands filtering, comparison analysis, event investigation, traffic channel and referrer analysis, and realtime analytics. For API consumers and self-hosted maintainers, unified contracts, generated documentation, compact internal site keys, event-driven ingestion scheduling, and stronger validation make the platform easier to extend and operate.

Highlights

• Added a unified typed Analytics Query architecture that provides shared query contracts, providers, caching, query cost controls, timeout handling, and cancellation handling across the Dashboard, API v1, SSR, Mock, and Realtime.
• Redesigned API v1 with unified request DTOs, response envelopes, route registries, permission declarations, and error handling, together with a generated API client. Added or expanded interfaces for comparison analysis, pages, referrers, traffic channels, filter values, performance, event and journey details, visitors, sessions, and realtime analytics.
• Added advanced filtering with nested AND / OR / NOT expressions, set filtering, ranges, null checks, and complex condition combinations. Added reusable saved filters with private and team-shared visibility.
• Added traffic channel analytics and referrer/channel trend charts for direct traffic, external referrals, organic search, social traffic, and campaign-tagged visits.
• Improved Dashboard tables, charts, filter panels, time tooltips, event details, and journey event details, making complex analytics pages easier to use while reducing unnecessary loading and rendering work.
• Improved Realtime data transport, event envelopes, channel subscriptions, realtime event details, and client Mocks so realtime pages and API responses remain consistent.
• Added a versioned traffic blocking-rule system for domains, paths, query parameters, referrers, User-Agents, IPs, ASNs, countries, and regions. Rules support Allow / Block behavior, wildcards, and match reasons while preserving compatibility with legacy configuration.
• Optimized D1 queries, data writes, map basemaps, and Durable Object ingestion scheduling to reduce runtime costs for high-traffic sites and long time-range queries.

Changes Since v0.6.0

• Migrated analytics queries from scattered query handlers into unified Contract, Application, Composition, and Provider layers. Query entry points now share input validation, query cost policies, result caching, provider error handling, and request-scoped observability.
• Reworked API v1 routes and data contracts. Existing paths such as /analytics/compare, top-level /events, /performance, and /realtime have been reorganized under the /analytics/* namespace, with additional site-level and team-level analytics interfaces.
• Added API v1 schemas, Comparison Analysis v2, batch requests, resource management, saved filters, site configuration, and generated client support. OpenAPI documentation is now generated from the Route Registry and Zod contracts and covers request, response, permission, and error definitions.
• Added a typed Filter Contract and versioned Filter DSL supporting condition composition, AND / OR / NOT, set operations, ranges, null checks, finite-set normalization, and JSON SQL parameterization for large sets.
• Added system filter presets and Filter Expression syntax help for common scenarios including direct traffic, external referrals, organic search, social traffic, campaign traffic, device types, and missing geographic attribution.
• Added the saved_filters table and a private/team-shared visibility model. Filter conditions are stored together with the user's original DSL and DSL version to support future compatibility and migration.
• Added page, referrer, and Traffic Channel Analytics. The Dashboard and API v1 can now report views, sessions, visitors, and trends by channel and support channel filter-value search.
• Unified Dashboard chart components and added implementations for latency percentiles, performance trends, request observation, event trends, share trends, traffic stacks, and dimension breakdowns.
• Improved analytics table and detail interactions, including column settings, infinite scrolling, lazy loading, detail drawers, event payload filtering, cross-page filters, tooltip presentation, and mobile scrolling behavior.
• Added a journey event detail drawer, enriched event, visitor, and session detail responses, and unified event-detail behavior across the Dashboard, Realtime Mocks, and API v1.
• Moved management-page and Demo queries to the server runtime, reducing SSR data requests and improving initial loading behavior for Team Dashboards and management pages.
• Applied broad Dashboard rendering optimizations, including memoization, component isolation, stable table props, deferred inactive panels, isolated map updates, stable chart point keys, and improved loading transitions.
• Migrated map basemaps from the legacy Tiles Handler to a Vector Relay, unifying basemap retrieval for geographic maps, realtime maps, visitor details, and session details.
• Reworked Realtime channel consumers and data transport to unify response structures for realtime snapshots, active visitors, events, sessions, and rolling trends.
• Optimized D1 query paths by reducing repeated scans and write amplification, removing redundant indexes without dedicated query paths, and improving event pagination, rollups, trends, and detail queries.
• Added site_identities and the site_pk internal-key migration. Historical data, events, visits, archives, and aggregation tables progressively switch from text Site IDs to compact integer keys while preserving the external site_id representation.
• Added event-driven alarm scheduling to the ingestion Durable Object, unifying flush due times, retries, backoff, timeout cleanup, and failure recovery.
• Replaced the Bot Analytics confidence model with threat levels and Request Observation categories, distinguishing high_threat, medium_threat, and custom_block.
• Changed Request Observation latency metrics to use sample-weighted aggregation for more accurate averages and weighted percentiles. The Dashboard primary metric now shows average latency, with p95 latency displayed in the details.
• Added parsing, validation, serialization, and matching for versioned Blocking Rules. Rules support Allow / Block behavior, wildcards, IP ranges, ASNs, query parameters, and geographic fields, and can return specific match reasons.
• Added a geographic search dialog for Blocking Rules and improved country, region, and city rule configuration.
• Improved English, Chinese, and Japanese copy for channels, filters, geographic locations, blocking rules, and Request Observation, and added an i18n watcher for development environments.

Developer Experience

• Added automatic generation of docs/schema.sql and consistency checks between the database schema and migrations.
• Extended OpenAPI and Skills Manifest validation. The Skills Manifest now includes credential guidance, permission descriptions, execution policies, error handling, and analytics task recipes.
• Expanded Agent task recipes from basic workflows to site inventory, traffic overviews, comparison analysis, filter exploration, performance analysis, event analysis, funnels, retention, visitor/session investigations, realtime monitoring, and site management.
• Improved the local Check Runner with task dependencies, parallel checks, execution progress, duration reporting, and Strict Verification Mode.
• Expanded test coverage for API v1, typed filters, saved filters, D1 site identities, Realtime, ingestion, Dashboard SSR, Blocking Rules, and Request Observation.
• Upgraded Cloudflare Vite Plugin, Wrangler, Workers Types, Hono, TanStack, React, and other dependencies while synchronizing generated API documentation and runtime contracts.

Upgrade Notes

• Some API v1 paths and response contracts have changed. Clients using the legacy /analytics/compare, top-level /events, /performance, or /realtime paths must migrate to the new /analytics/* paths and update request and response handling according to the latest OpenAPI documentation.
• API keys used for saved filters or shared Analysis Definitions must have the corresponding Analysis Read / Write permissions. Site discovery, site configuration, and Analytics Queries remain controlled by their respective API scopes.
• Existing Blocking Rule configuration continues to work through the compatibility layer. No manual migration is required when upgrading unless users want to adopt the new wildcard or Allow / Block rule capabilities.