#Inspiration Index

• Automatic collision detection that triggers S3 auto-processing and simplifies the retry workflow
• Pipeline now consolidates page type taxonomy into a single source of truth for consistent classification across the system
• Fresh sidebar navigation and enhanced pipeline review UI, plus a complete theme system migration
• Link scraper adds intelligent page discovery and classification—automatically detecting and categorizing page types as it crawls
• Page type hints passed directly to the Recorder API; metadata batched per-URL for better tracking

#Autoscroll Recorder

The big win: unlimited distribution retries with timeout detection. Jobs that fail to distribute now automatically retry with detailed technical tracking. The distribution system displays real-time status on job detail pages, making it easy to monitor what's happening with your recordings.

• Improved S3 destination setup flow with auto-generated external IDs and a corrected workflow
• Migrated to JWT authentication and Terragrunt infrastructure
• Support for optional per-job metadata via the API
• iubenda cookie consent banner support
• Comprehensive S3 metadata with every recording
• Lambda now properly persists user metadata to the database

#PostgreSDK

Breaking change: multi-service authentication. Replaced the single-secret JWT system with proper multi-service auth, making it production-ready for distributed systems.

• Added an init pull subcommand for simpler configuration
• Increased max pagination to 1,000 records
• Direct nested includes now work without wrapper syntax
• JWT secrets no longer hardcoded in generated code
• CLI shows helpful errors when config files are missing
• Documentation overhauled with new guides for service authorization and SDK distribution

#Agentic Editor

Migrated infrastructure to use shared VPC resources, reducing operational complexity and improving resource utilization across environments.