Interview Questions

Interview Question Generator

What This Skill Does

Produces a structured interview question bank tailored to a role, seniority level, and the specific traits being assessed. Crucially, each question comes with what a strong answer vs. a weak answer looks like — so the interviewer isn't just asking questions, they know what signal to listen for. This keeps a panel aligned and stops interviews from becoming unstructured chats.

---

Process

1. Get the inputs: role, seniority, and the 3–5 traits/competencies being hired for (pull from

the rubric if one exists). Ask if missing.

1. Generate questions in three buckets:

• Behavioral — past behavior as predictor ("Tell me about a time...")
• Competency / technical — can they actually do the work
• Role-specific / situational — judgment in scenarios they'll actually face

1. For each question, write the listen-for: what a 5-answer sounds like, what a 1-answer sounds

like. This is the part that makes the guide useful.

1. Tie questions to rubric dimensions where one exists, so scoring is clean.
2. Include probes — follow-ups that dig past a rehearsed first answer.

---

How to Write Good Questions

• Behavioral beats hypothetical. "Tell me about a time you shipped something nobody asked for"

surfaces real slope. "What would you do if..." invites a rehearsed ideal answer.

• Make questions surface the rubric signals: quality of thinking (ask open questions and watch

what they ask back), slope, self-awareness (poke holes in their answers and watch them adapt).

• Tie at least one question to self-awareness under pushback — present something they did and

challenge it; the adaptation is the signal.

• Avoid trivia and gotchas. They test memory and nerves, not the job.
• Calibrate to seniority. A senior hire should face ambiguity and tradeoff questions; a junior

hire, fundamentals and learning speed.

---

Output Format

# Interview Guide: [Role] — [Seniority]
**Assessing:** [the traits/competencies]

## Behavioral
**Q:** [question]
→ *Strong answer:* [what good sounds like]
→ *Weak answer:* [what a red flag sounds like]
→ *Probe:* [follow-up]
*(Maps to: [rubric dimension])*

## Competency / Technical
[same structure]

## Role-Specific / Situational
[same structure]

## Open Floor (Quality-of-Questions test)
[Reminder to leave space for the candidate's questions and treat what they ask as data.]

---

Question Failure Checklist

A guide FAILS if it:

• ❌ Asks questions with no "listen-for" — leaving interviewers to guess what's good
• ❌ Leans on hypotheticals where behavioral questions would surface real evidence
• ❌ Uses trivia/gotchas that test nerves, not the job
• ❌ Isn't calibrated to seniority
• ❌ Doesn't surface the rubric's key signals (slope, self-awareness, quality of thinking)
• ❌ Has no probes, so a rehearsed first answer ends the thread

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: getclera
• Source: getclera/claude-skills
• License: MIT

Install and usage instructions live in the source repository linked above.