thinkhard.SI

Prompt repository Hiring & people

Turn rough interview notes into evidence-based feedback

You have just finished an interview and need to write up feedback for the hiring team from scribbled notes, tied to the scorecard and free of snap judgments.

Registry no.
PPL-003
Collection
Hiring & people
Laid out for
Any chat model
Score
10 out of 10
Blanks to fill
3
Added

Hiring & peopleAny chat

The prompt

Score 10 of 10, Excellent

Use it when You have just finished an interview and need to write up feedback for the hiring team from scribbled notes, tied to the scorecard and free of snap judgments.

  • {{role_title}}
  • {{scorecard}}
  • {{interview_notes}}
Full prompt and why it works
Prompt
Turn my rough notes from an interview into written feedback for the hiring team for the {{role_title}} role. The team will compare feedback from several interviewers before a person makes the decision, so the write-up has to show evidence rather than my overall impression.

Competencies I was asked to assess, with the scoring guide:
"""
{{scorecard}}
"""

My notes from the interview:
"""
{{interview_notes}}
"""

Write the feedback in this format:
- One section per competency, with three labeled lines: Evidence (what the candidate said or did, from my notes, as specific as the notes allow), Gaps (what I did not see or did not ask about), Suggested score (from the scoring guide, with one sentence on why).
- Then "Overall": two or three sentences on strengths and concerns, based only on the sections above.
- Then "Follow up in a later interview": up to 3 questions another interviewer could ask to fill the gaps.

Keep it under 350 words. Write neutral, factual sentences, such as "Described leading the migration of the billing system and named the rollback plan they wrote", rather than labels like "great energy" or "seemed smart". The feedback should read as a record of what happened in the room, so that a colleague who was not there can see why each score was suggested.

Rules:
- Use only what is in my notes. Do not add details, quotes or scores the notes do not support.
- Leave out anything about the candidate's age, family, health, accent, appearance, religion, nationality or any other protected characteristic, even if it appears in my notes, and list it under a final line "Removed from notes" so I can see what was taken out.
- Replace vague impressions such as "not a culture fit" with the specific behavior behind them, or mark them "unsupported" if the notes give no behavior.
- Do not recommend hire or no hire. The suggested scores are mine to confirm, and the decision belongs to the hiring team.

If my notes are too thin to score a competency, write "Not enough evidence" as its score; do not guess.

Variables

{{role_title}}
The job title.
{{scorecard}}
The competencies you were asked to assess and the scoring guide for each, from the interview plan.
{{interview_notes}}
Your notes from the interview, as rough as they are.

Why it works

  • What to avoid. Protected characteristics are removed even when they appear in the notes, and listed, so the interviewer can see what was taken out.
  • What good looks like. Labels like "not a culture fit" must be replaced with the behavior behind them or marked unsupported, which is what makes feedback comparable across interviewers.
  • If unsure. Thin notes produce "Not enough evidence" instead of a guessed score, so gaps stay visible to the hiring team.

How to use it

  1. Press Fill in variables and type your details into each blank. What you type stays in this browser.
  2. Copy the filled prompt into the AI tool you already use, such as ChatGPT, Claude or Gemini, or open it in the Grader or the Prompt Tester here.
  3. Before you rely on it, try it on three or four real examples. The Prompt Tester runs them all at once and shows which ones go wrong.
Use it from code

Get this prompt as JSON from the free REST API, with your own text in the blanks, or let an AI assistant fetch it through the MCP server. No key.

curl -s https://thinkhard.si/api/v1/prompts/interview-notes-to-scorecard-feedback

With the MCP server added to Claude, ChatGPT, Cursor or VS Code, call get_prompt with {"prompt": "interview-notes-to-scorecard-feedback"}, or just ask for it by name.