Discovering is not publishing
It began with the daily task of tracking news in games, anime and pop culture. With so many sources, it is hard to distinguish a story worth covering from a repeated announcement or something already published on the site.
The system monitors official sources and signals of relevance. Other outlets can show that a topic is gaining attention, but we seek primary confirmation whenever possible. Gathering candidates does not decide what gets published.
A workflow with editorial criteria
Before drafting, the bot compares URLs, titles, topics and recent posts to reduce duplication. It then scores relevance based on the source, timeliness, impact and fit for readers. A weak story stops there.
With enough reliable information, AI helps draft an original article. Editorial rules guide the language, tone, sources, images, links, category and format. The workflow checks and cleans the material before sending it to WordPress.
Automation helps discover and prepare stories. Editorial judgment determines what gets published.
Reviewing is part of the system
The dashboard separates story candidates, drafts, review and publication. The cautious workflow prepares material for human review. Direct publication requires explicit settings and operational validation, with specific criteria for sources and images.
The architecture makes one principle clear: asking AI to rewrite news is not enough. We need to know where information comes from, what is confirmed, what remains speculation and whether the article adds context for readers.