Every AI assistant has the same tic: three paragraphs of throat-clearing before it tells you the thing you actually asked. “Let me look into that… I should also mention… here’s a summary of what I found…” By the time the answer shows up, you’ve lost the thread.
ayghri/i-have-adhd is a Claude Code skill that just turns that behavior off. Once it’s active it rewrites how the model talks:
- Answer or next action comes first, always
- Multi-step work gets numbered, one bounded action per step
- No preamble, no recap, no closing summary
- Progress carries across turns instead of resetting each time
- Time estimates are concrete, not “soon” or “shortly”
It was built for ADHD accessibility, but it’s picked up thousands of installs from people who just want tools that don’t bury the lede. Turns out most of us have the same threshold for wading through filler.
Installing it takes two commands:
claude plugin marketplace add ayghri/i-have-adhd
claude plugin install i-have-adhd@i-have-adhd
Then /i-have-adhd once per session, and it stays on for the rest of the conversation.