An AI visual feed idea would turn character dossiers into ComfyUI-generated posts
tovarischsht · reddit · 2026-07-28
A Reddit user proposes an AI-generated visual feed that would turn character dossiers and LoRA presets into pseudo-Instagram posts.
The workflow would be:
- define each character as a dossier with personality, location, habits, and history
- attach a LoRA and trigger words
- auto-generate fake social posts and prompts
- send those prompts into a predefined ComfyUI workflow
- store the generated posts in a database and browse them through a feed-like viewer
The appeal is to skip manual prompting when the goal is just to browse random visuals, while still allowing feedback loops through likes and comments. The idea could also be extended beyond characters to nature shots, art, or even RSS/news aggregation. The poster notes that existing projects like feedme could be forked to fit this model.
More from Multimodal
- Google demos Gemini 3.5 Flash-Lite on 1M+ catalog images for structured extraction — gaganghotra_ · 2026-07-28
- Topview Film Studio turns AI video generation into a cinematic directing canvas — eyishazyer · 2026-07-28
- Flux 3 reportedly nails a National Geographic-style video from a single prompt — FinanceYF5 · 2026-07-28
- Depth Anything V2 motion maps survive character swaps in a Seedance 2.0 test — Few-Profession421 · 2026-07-28
- UniFlex 13 packages free Krea 2 ComfyUI workflows with one-node setup — kaptainkory · 2026-07-28
- Grok prompt asks for a Wednesday character in Barbie-movie style — heypearlai · 2026-07-28