Changelog
All notable changes to FacilFlow are documented here.
[0.3.0] - 2026-01-25
Added
- InfluxDB 3 Clusters: Docker-based cluster provisioning with multi-node topologies (querier, ingester, compactor, router), container lifecycle management, live log streaming, and built-in data explorer with SQL
- ML Models: MLflow integration with experiments, runs, model registry, staging workflow (None/Staging/Production/Archived), approval workflows, and inference pipeline nodes
- Schemas: Schema versioning with SHA-256 checksums, compatibility checking (backward, forward, full, none), version history, rollback, and side-by-side comparison
- Dashboards: Widget-based dashboard builder with Chart, Table, and Metric widgets, grid layout, sharing with view/edit permissions, and duplication
- Data Quality: Rule-based quality evaluation across five dimensions (completeness, accuracy, consistency, validity, timeliness), quality scoring, trend monitoring, and quality alerts
- Data Lineage: Automatic lineage graph with upstream/downstream tracking, pipeline lineage capture, React Flow visualization, and impact analysis
- Alerts: Alert rules with conditions, multi-channel notifications (Slack, Email, PagerDuty, Webhook, Microsoft Teams), escalation policies, alert history with acknowledge/resolve/silence, and cooldown
- Scheduler: Cron-based and event-based pipeline scheduling, visual cron expression builder, execution history, and job queue
- Connectors Hub: Unified connector catalog with Airbyte (300+ sources), Telegraf plugins, public API repository, and custom connector support
- Governance: Data classification levels (Public, Internal, Confidential, Restricted, PII), resource tagging with key-value pairs, audit logs, and tenant isolation
- Feature Engineering: Pipeline node for computing derived features (rolling stats, lag, ratios, time-based)
- Pipeline Templates: Pre-built templates for common patterns (OPC-UA to InfluxDB, multi-source join, ML scoring, quality monitoring)
- Pipeline Test Mode: Validate pipelines with sample data before deploying
- Pipeline Export/Import: Download and upload pipeline configurations as JSON
Changed
- UI migrated to shadcn/ui: Replaced MUI with shadcn/ui, Radix primitives, and Tailwind CSS
- 5-section navigation: Reorganized sidebar into Connect, Infrastructure, Build, Analyze, and Operate sections
- Pipeline builder: Added Join, Dataset, Feature Engineer, and ML Inference node types; undo/redo; keyboard shortcuts
- Edge Agents: Added Telegraf config generation, agent-to-cluster mapping, remote configuration push
- Data Catalog renamed to Data Sources: Expanded with connection testing, multi-protocol support, CRUD operations
[0.2.1] - 2026-01-15
Added
- Jarvis Chat Panel: Floating assistant with real-time SSE streaming responses
- Artifact Rendering: Display charts, tables, code blocks, and Mermaid diagrams from AI responses
- Canvas Panel: Side panel for viewing and switching between generated artifacts
- Claude Code Integration: Backend uses Claude Code CLI for OAuth-based authentication
Changed
- FloatingAssistant now connects to backend API with streaming support
- Chat messages support embedded artifacts for rich AI responses
[0.2.0] - 2026-01-13
Added
- MinIO Object Storage: File storage integration with S3-compatible MinIO backend
- File Upload Component: Drag-and-drop file uploads with progress indication
- Storage API: REST endpoints for file upload, download, and management with presigned URLs
- Real-time Pipeline Monitoring: Live status updates for pipelines via WebSocket connections
- Real-time Agent Monitoring: Live status and heartbeat tracking for edge agents
- Jarvis AI Assistant: Natural language interface for pipeline management powered by Claude
- Pipeline Tools: AI can list, create, start, stop, and delete pipelines via chat
- Query Tools: AI can explore data and run queries via natural language
- Chat History: Persistent conversation storage across sessions
- Data Connectors: OPC-UA, PI System, and Airbyte source node integrations
- Airbyte Source Node: Visual pipeline node for 300+ data source connectors
- Database Production Ready: Migrations, health checks, and seeding utilities
- Data Sources Service: CRUD operations with connection testing for multiple protocols
Changed
- Backend migrated to full TypeScript with Drizzle ORM
- Health endpoints now include database latency metrics
- Improved graceful shutdown with proper connection cleanup
[0.1.0] - 2026-01-12
Added
- Visual Pipeline Builder: Drag-and-drop interface for creating data pipelines
- Custom Pipeline Nodes: Source, Processor, Sink, and Edge Agent node types
- Edge Agent Node: Connect pipelines to remote edge agents
- Node Palette: Categorized node library with search
- Authentication: Firebase-based login with email/password and Google SSO
- Role-Based Access Control: Admin, Operator, and Viewer roles with granular permissions
- Protected Routes: Route guards based on authentication and permissions
- Agent Management: List, view, and manage edge agents with status indicators
- Data Explorer: Query and visualize time-series data from InfluxDB
- Time Range Selector: Flexible time range picker for data queries
- SQL Query Builder: Write and execute SQL queries against InfluxDB 3
- Results Table: Paginated data display with sorting
- Data Charts: Visualization of query results
- InfluxDB 3 Integration: Native SQL queries with Arrow Flight streaming
- Dark/Light Theme: Toggle between light and dark modes
- Responsive Sidebar: Collapsible navigation with FacilEdge-style design
- E2E Tests: Playwright tests for authentication and pipeline builder flows