Checks robots.txt line by line and says what each AI crawler token actually controls - training, AI-search citation, or user fetch - and which lines silently do nothing. (bundled Node.js runtime)