Interview scorecard

One row per competency, one score from 1 to 4, and a column for the evidence you actually heard. What each score means is printed on the card itself, so two interviewers reading the same answer are at least using the same yardstick, and a score with no evidence next to it is visibly incomplete.

Four levels have no midpoint, so the card cannot be filled in entirely with the middle score. Choosing between 2 and 3 forces a position: either there is a gap you would have to cover after the hire, or there is evidence they have already done this at the level the role needs.

From the must-have list in your own job description, which is why the JD checklist is linked from this page. Type them in and they appear on the card unchanged. The ten fixed lists, and the two questions under each competency, are editorial starting points written by hand, not generated, and you are expected to edit them. Recognized competencies use fixed role-specific questions and rating guidance. Other competencies use a general behavioral template to adapt before the interview.

No. It reads no resume, takes no candidate data, and makes no prediction. It lays out the same questions and the same scale for every candidate in one role; the scoring is done by the people who sat in the interview.

The card and share link are built in your browser; field contents are not sent for generation. Interaction analytics may send usage events. The optional email form sends its fields only when you submit it.

Yes. Copy as Markdown gives you a table to paste into a doc. Copy link to this card writes ?role=, ?stage=, and ?competencies= into the query string. Opening it fills the card in the browser and does not call a server. Ten competencies is the maximum. Neither one carries a candidate name, because the page never asks for one.