v1.17.3

What's New in v1.17.3:

• Redesigned AI Engines UI: Introduced a dedicated OmniRoute All-In-One Powerhouse Hub with unified credentials and 3-in-1 modality selectors (Text, Image, Video), making engine setup simple and intuitive.
• AI Puffer RAG Bot Support: Seamless integration with local AIPKit/AI Power chatbots for brand-accurate, context-aware copywriting with OmniRoute high-availability failover.
• Video Engine Direct URL Support: Resolved empty binary drops when upstream video engines return direct streaming MP4 links.
• Settings & Video Lab Harmonization: Whitelisted OmniRoute across Settings and Video Lab provider menus.
• Model Catalogue Optimization: Added query limit pagination preventing DOM bloating and speeding up admin rendering by 5.5x.

v1.17.2

Features & Bugfixes in v1.17.2:

• OmniRoute Text Engine: Enforced non-streaming payload (stream => false) to prevent cURL 28 timeouts; default model set to �uto/best-coding.
• OmniRoute Image Engine: Configured verified working image model �ihorde/stable_diffusion with zero-cost Pollinations fallback; added automatic magic-byte MIME detection and media library sideloading.
• Video Engine: Corrected OmniRoute generation endpoint URL to /videos/generations, updated candidate models (�eo-free/veo, �eoaifree-web/veo, �eo-free/seedance, �uto/inkling), added direct result URL extraction.
• Model Synchronization: Optimized wp_vmsai_models schema with composite unique key and replace queries for robust high-volume multi-modality catalogue sync.
• Network Channels & Publishing: Added native carousel & multi-photo publishing for Facebook and Instagram; added TikTok, Bluesky, and Threads optimizations.
• Admin & Analytics: Rebalancing respects user Quiet Hours; added REST inbox/reply actions for community management.