# Crawl / LM queue weights by LLM category (lowercase exact labels). # Higher = fetched and summarized sooner. Unlisted categories use `default`. # Restart fetcher + worker after edits. Links are enqueued only after LM classifies # the page that contained them — not dropped, just ordered. Language demotion is # in language-priority.txt (non-en / mul) and added on top of these weights. # # Goal: people making things for a community (art, code, hardware, writing), # not corporate products. The model usually dumps those sites into personal/blog; # the specific labels below are rarer and higher precision — boost them hardest. # # seed = ingest / spider seed hosts # default = unknown / unlisted category # negative = still crawled eventually, after everything better seed 50 default 0 # makers — a person or small crew shipping a thing portfolio 40 art-portfolio 40 artistic-portfolio 40 game-portfolio 40 art 35 hobby 35 art-gallery 30 gallery 30 comics 30 comic 30 literary-journal 30 literature 28 community 28 podcast 25 music 25 music-band 25 software 24 magazine 22 music-blog 22 wiki 22 # forumotion-style spam boards were riding +22 into B2B mills forum -12 radio 20 tools 18 tech-blog 18 science-blog 18 fanfiction 16 fan-fiction 16 game 16 gaming 14 technology 12 fandom 12 archive 10 library 12 museum 8 nonprofit 8 non-profit 8 charity 8 zine 30 arts 28 # people-shaped but mixed (most neocities land in personal; tumblr in blog) blog 12 personal 6 other 0 # fandom directories / shrines — useful graph later, not the main hunt # fan-site was +20; that dumped ~9k tumblr + a 38k pending pile. Enough. fan-site 0 entertainment 0 # demote — still indexed, outbound trickle education -5 academic -5 academic-journal -5 government -8 organization -8 politics -8 political -8 news -12 tech-news -10 media -12 sports -10 sports-news -10 social-media -15 saas -18 documentation -18 ecommerce -20 retail -20 corporate -25 parked -30 empty -30 gambling -40 gambling-site -40 authentication -25 finance -20 financial-service -20 financial-services -20 real-estate -20 hotel -20 hotel-booking -20 travel -15 tourism -15 travel-agency -15 travel-guide -12 travel-booking -20 restaurant -15 healthcare -15 health -12 health-information -12 legal -15 legal-services -15 directory -12 search-engine -12 consulting -15 streaming-service -18 job-board -15 jobs -15 dating -20