ARMIYA-CTRL/ai-prop-protection.com

This is a 0-star, 200-day-old repository with no adoption signals (no forks, no velocity). The core idea—detecting propaganda in AI outputs—is not novel; it applies straightforward text analysis heuristics (likely keyword matching, bias word lists, or basic NLP) to a new context (browser extension overlay). The implementation appears to be a prototype-stage proof-of-concept rather than production-grade. No evidence of active maintenance, community adoption, or technical depth. Frontier labs (OpenAI, Anthropic, Google) have no incentive to build this as a standalone tool; they're more likely to integrate basic safety/bias-detection signals directly into their platforms (as they already do with content policies and system prompts). The extension sits in a crowded space of fact-checking and bias-detection tools without apparent differentiation. Defensibility is minimal—the pattern-matching logic is trivial to replicate, and no data advantage or switching cost exists. User retention depends entirely on extension adoption friction, which is high. Low frontier risk because frontier labs solve this problem within their own products and would not acquire a niche browser extension.