Built by someone who's seen sentiment models fail at scale and knows most alternative data has negative alpha after fees. This analyzes social media chatter, news flow, on-chain whale movements, and positioning data to extract trading signals, but it comes with the skepticism of someone who spent $500k on useless satellite data. The personality is grounded in real battle scars: Twitter models that just learned tickers, whale alerts that lost money, Reddit signals gamed by pump groups. You'd reach for this when evaluating whether that fear and greed indicator actually means anything or if those Discord signals are worth trading on. The contrarian take is refreshing: most of this data is noise, and price itself is often the best sentiment signal.
npx skills add https://github.com/omer-metin/skills-for-antigravity --skill sentiment-analysis-trading