f77226532e
Backend: - When content_types includes "image", generate image via configured image provider (DALL-E, Replicate) after text generation - Image prompt built from character niche + topic + visual style - Graceful fallback: if image generation fails, post is text-only - Post content_type set to "text+image" when image is attached Frontend: - ContentPage preview shows generated image above text when present - Image displayed with full width and border Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Leopost Full
Social Content OS — merge di Leopost, Post Generator e Autopilot OS.
Lab: https://lab.mlhub.it/leopost-full/
Stack
- Frontend: React (Vite) + Tailwind CSS — design system "Editorial Fresh" (Fraunces + DM Sans)
- Backend: FastAPI (Python) + SQLite
- AI: Multi-LLM (Claude / OpenAI / Gemini)
- Publishing: Facebook, Instagram, YouTube, TikTok
Funzionalità
- Gestione Characters (profili creatori)
- Generazione contenuti AI multi-piattaforma
- Calendario editoriale con livelli di consapevolezza (PAS, AIDA, BAB...)
- Export CSV per Canva Bulk Create
- Scheduling automatico + pubblicazione social
- Gestione commenti con AI reply
- Affiliate links con auto-injection
- Multi-LLM configurabile via Settings