free, no account
Everyone says they use AI well. Here is a second opinion.
Paste a real conversation you had with an AI assistant. You get scored on the four things that separate directing an assistant from hoping it guesses right, with the evidence quoted from your own words.
What you are scored on
- Understood the task. The learner framed the goal for the assistant clearly instead of pasting the brief and hoping.
- Grounded in the material. The learner directed the assistant into the provided files and based the work on them, not on invented facts.
- Verified the output. The learner checked claims, numbers, or coverage against the source material before submitting.
- Iterated with judgment. The learner refined weak parts of the draft with specific follow-ups rather than accepting the first answer.
These are the same criteria every DailyByte mission is graded on, not a simplified version of them.
200 more characters needed. A few real exchanges is enough.
Your conversation is sent to a model for grading and is not stored. What is saved is the scores and the short lines the grader quotes as evidence, on an unlisted page you can delete.
Common questions
Do you keep my conversation?
No. It is sent to the grading model and not written to our database. What is stored is the scores, the headline, and the short evidence lines the grader quotes, which is the least that can be kept while the result page still says anything useful. Your result page is unlisted and you can delete it.
What counts as a good score?
Most people lose marks on verification: they check whether an answer reads well rather than whether it is true against something. A high score means you framed the task, pointed the assistant at real material, checked what came back, and pushed on the weak parts rather than accepting the first draft.
Is this the actual product?
It is one half of it. DailyByte scores the same way on realistic work missions for your role, and also grades the artifact you produce. See the role paths for what that looks like.