Welcome Daily.dev and Reddit developers! Share your agent use cases here.

[UditAkhourii]

Ask your questions, show your work and tell me more about what you are building..

[UditAkhourii]

Thanks for the welcome. This is exactly the kind of conversation I want happening here.

If you are testing ADHD in your projects, the most useful thing you can share is one specific moment: the prompt you ran, the answer the baseline gave, and the answer ADHD gave. Even when ADHD lost, that data point is worth more to me than a dozen "it works great" comments.

Three things I am especially curious about right now:

Cases where the regulator, biology, or 10-year-old frame surfaced an answer no other frame got close to.
Cases where ADHD over-spent tokens on a problem that did not actually need the wide search, so I can tighten the pre-flight gate.
Use inside other people's agents at decision points (planning, code review, refactor branch points). repowire just shipped ADHD in their orchestrator template (PR #313).
Drop a comment, open an issue at https://github.com/UditAkhourii/adhd/issues, or sign up to the community at https://tally.so/r/WO1Nzj if you want to join as a contributor, maintainer, early adopter, or just a member.

Building this in public, including the parts where it does not work. Looking forward to your stories.