If you mostly use ChatGPT or Claude, the easy mistake is to see ayghri / i-have-adhd and file it as niche. Wrong. The useful idea is broader: it got traction for cutting AI padding, not for the label. Skip the warm-up. Give the answer. Then the next move. [C002]
A lot of people are tired of asking a small question and getting four polite paragraphs before one usable sentence. If you only use chat-style AI, that pain is not niche. It is the daily tax of padded answers.
The repo page for ayghri / i-have-adhd [C001] showed 6.5k stars and 259 forks when I checked. That does not prove real-world results. It does suggest the pain travels well: plenty of users want the answer before the ceremony.
The best proof is the before/after example. It cuts lines like "Great question" and "Hope this helps," then keeps the part that changes your next move: the action, file path, numbered steps, and what to do next. That is anti-padding, not anti-human.
My filter for new AI tools: do not ask how smart the pitch sounds. Ask whether it changes your next action. I only checked the GitHub page and setup notes, not live results. Share this with the person who keeps getting polite fluff instead of an answer.