What the history actually shows
Ask for a look back and the agent calls get_my_interview_history, which returns every past session: what type it was, general, technical, resume-based, salary negotiation, comprehensive, or job-description-based, the role and date, whether it was completed, your scores across the five dimensions, overall, communication, technical, behavioral, and cultural fit, and the feedback highlights tied to each, both strengths and areas to improve.
From there the agent's value is in the read, not just the retrieval: it can point out that your communication score has climbed over your last three sessions while technical has stayed flat, or that three separate pieces of feedback are all circling the same habit worded differently.
If a gap looks worth practicing directly, it can offer to set up a new practice interview focused on that specific area with create_practice_interview, rather than leaving you to guess what to work on next.
Finding what keeps coming up
You ask: "how did my last few practice interviews go, and what feedback keeps coming up?" The agent calls get_my_interview_history and looks across your last four sessions.
It tells you your overall and technical scores have both been solid and steady, but the behavioral feedback in three of the four notices the same thing: your answers tend to skip the outcome and stop at what you did, without saying what actually resulted from it.
That's specific enough to act on. You ask it to set up a practice session focused on that. It calls create_practice_interview with a behavioral focus and hands you a link to start it right away.
Nothing about reading your history or noticing the pattern required approval, since it's all read-only; setting up the new practice session does, and the card names exactly what kind of interview is about to be created before it exists.
Boundaries
What stays scoped to you
Pulling up your interview history doesn't touch an approval card at all, since reading it changes nothing. Every session, score, and piece of feedback returned belongs to you specifically; the tool has no path to another candidate's interviews, practice or otherwise, and it can't be asked to compare you directly against someone else's results.
The scores it reports are the same five-dimension scores, overall, communication, technical, behavioral, cultural fit, generated at the time each interview was actually graded; the agent doesn't recalculate or adjust them when summarizing your history, it reports what already exists.
Setting up a new practice interview based on what it notices is the one write in this flow, and that step asks your approval and names the interview type before creating anything. Looking back at your own history costs nothing, the same as the rest of the agent.
Frequently asked questions
Can the agent tell me if the same feedback keeps showing up across my interviews?
Yes. It reads your full interview history at once, scores and feedback for every session, and can point out a pattern that repeats across more than one, something harder to notice reading one report at a time.
Does the agent see anyone else's interview results?
No. Every session it reads back is scoped to your own account. There's no path from this tool to another candidate's interview history or scores.
What scores does the history actually include?
The same five dimensions each interview is graded on: overall, communication, technical, behavioral, and cultural fit, along with the written feedback tied to each completed session.
Can it set up a new practice interview based on what it notices?
Yes, if you want it to. It can offer a practice session focused on a specific gap it noticed, and setting one up requires your approval and names the interview type before it's created.
Related pages
See what your interviews are telling you
Ask the agent to look back across your practice interviews, no subscription required for this part. It only ever reads your own results, nobody else's.