# Interview Prep

> Use to prepare for a specific interview, when a user says "I have an interview", "prep me for the [company] screen", "help me get ready for the hiring-manager round", or names an interview stage. Reads the JD, company research, and story bank, builds a stage-specific prep pack (likely questions, mapped stories, positioning, objections, questions to ask), and optionally runs a mock interview with rubric critique. Before the interview only. Afterwards use `interview-followup`, or `interview-debrief` if a transcript exists.

- Skill: `guerrilla2799/interview-prep` (Agent Skill)
- Install (CLI): `npx skillmds@latest add guerrilla2799/interview-prep`
- Raw SKILL.md: https://api.skillmd.com/api/skills/guerrilla2799/interview-prep/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: guerrilla2799 (https://skillmd.com/u/guerrilla2799)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/guerrilla2799/interview-prep

---


# Interview Prep

Turns "I have an interview Thursday" into a concrete prep pack tailored to the exact role, stage, and people in the room, then drills it with a mock. Prep to the specific stage you're in. Each stage has a different bar.

## When to use
- "I have a [recruiter screen / hiring-manager / panel / technical / final] interview"
- "Prep me for [company]", "what will they ask me", "run a mock interview"
- After scheduling an interview and before the day of, ideally 2–3 days out so gaps can be filled

## Inputs
- Reads:
  - The **JD** for the role (from `pipeline/applications/<id>-...md` or pasted)
  - `workspace/output/<id>-company-research.md` (from `company-research`)
  - `workspace/stories/story-bank.md` (from `story-bank`)
  - `workspace/profile/profile.md` (positioning, constraints)
  - The application record `pipeline/applications/<id>-...md` (the role thesis, known contacts)
- Needs from user: the **stage** and **format**, the **interviewer names/titles** if known, the date, and anything the recruiter told them about the process. If company research or a story bank doesn't exist yet, say so and recommend running those first. Prep is weak without them.

## Stage map (each has different questions and a different bar)
| Stage | What they're really testing | Prep emphasis |
|-------|-----------------------------|---------------|
| **Recruiter screen** | Baseline fit, comp alignment, communication, no red flags | Tight 90-sec story, why-this-role, comp range answer, no rambling |
| **Hiring manager** | Can you do the job, and will they enjoy managing you | Role thesis, 2–3 deep STAR stories, how you'd approach their actual problems |
| **Panel / team** | Cross-functional fit, breadth, collaboration | A different story per interviewer's function, with consistency across the room |
| **Technical / case** | Actual skill under pressure, and how you think | Domain drills, a structured problem-solving framework, thinking out loud |
| **Executive / final** | Strategic altitude, judgment, culture, closing | Vision-level answers, sharp questions for a leader, why-you-vs-alternatives |

## Workflow
1. **Identify stage + format.** Confirm which stage this is and the format (call/video/onsite, 1:1/panel/case). State it back. The whole pack keys off this. If the user doesn't know who's in the room, note it as a gap and prep the most likely composition for the stage.
2. **Build the likely-question set for THIS role and stage.** Not a generic list. Derive questions from three sources: the JD's stated responsibilities and requirements (each one is a probable question), the stage's testing focus (table above), and anything company research surfaced (a recent reorg, a public struggle, a strategic bet). Produce 12–20 questions the user is genuinely likely to face.
3. **Map each behavioral question to a story-bank story.** For every behavioral question, name the specific story from `story-bank.md` that answers it best and note the one-line angle to lead with. If a likely question has **no story to cover it**, flag it loudly. That's the highest-value thing to fix before the interview, and it's a prompt to run `story-bank` to fill the gap.
4. **Write the role thesis / positioning.** 2–3 sentences: why this candidate for this specific role, the proof points behind it, and the through-line they should return to. This is what "walk me through your background" and "why you" should always land on.
5. **List the objections about their fit, and the answer to each.** Be honest and specific about the gap the resume shows (missing a named tool, a shorter tenure, a different industry, a level jump, an employment gap). For each, write the crisp, non-defensive reframe. Interviewers WILL probe these. Rehearsed beats surprised.
6. **Company-specific talking points.** Pull 3–5 concrete, current facts from company research (recent funding, a product launch, a strategic pivot, a competitor pressure) and phrase how the user connects to each. Evidence they did the work, not flattery.
7. **Write 5–8 questions to ask THEM, mapped to who's in the room.** Different questions land with a recruiter (process, comp, timeline), a hiring manager (the role's real problems, what success looks like at 90 days, why the seat is open), a peer panel (how the team actually works, what's hard here), an exec (strategy, where the bet is, how they'd know this hire worked). Tag each question with who to ask it. Avoid anything answerable from the website.
8. **Offer a mock interview.** If the user wants it, run one: ask questions **one at a time**, wait for the answer, then critique against the rubric below before moving on. Stay in role. Escalate difficulty toward the real stage's bar. Keep it to 5–8 questions unless they want more.
9. **Save the pack** to `output/<id>-interview-prep.md` and update the application record's Interview-prep section with a pointer, the interviewer names, and the format.

## Mock-interview critique rubric
Score each answer 1–5 on four axes and give one concrete fix per axis that's weak:
- **Structure.** Clear STAR (or a clear framework for a case), no meandering, a point you can follow
- **Specificity.** Real details, names, and numbers, not a generic "I collaborated cross-functionally"
- **Outcome.** Lands on a quantified or concrete result and answers "so what happened?"
- **Concision.** Under ~2 minutes, leading with the answer rather than a five-minute wind-up

Call out filler, hedging, buried leads, and "we" where it should be "I". Praise what genuinely worked. Don't only critique.

## Output
- Writes to: `workspace/output/<id>-interview-prep.md`
- Format:
  - `## Stage & format` · `## Interviewers` (names, titles, likely focus)
  - `## Role thesis` (the 2–3 sentence through-line)
  - `## Likely questions` – each with the mapped story-bank story (or a **GAP** flag) and lead angle
  - `## Objections & answers` – the fit concerns and the reframe for each
  - `## Company talking points` – 3–5 current facts + your connection to each
  - `## Questions to ask them` – 5–8, each tagged with who to ask
  - `## Mock log` (if run): questions, answers, rubric scores, fixes

## Rules & quality bar
- **Prep to the stage.** A recruiter-screen pack and an exec-round pack should look different. Generic prep is the failure mode.
- **Every behavioral question maps to a real story or is flagged as a gap.** No hand-waving. If there's no story, say so and route to `story-bank`.
- **Objections are named, not hidden.** The gaps in the fit are exactly what gets probed. Surface them here so the user isn't caught flat. This mirrors the OS-wide rule that gaps stay visible.
- **Questions-to-ask must be un-Googleable.** If the answer is on the About page, cut it.
- **No fabrication.** Company facts come from research and must be current and true. Stories come from the bank. Don't coach the user to claim experience they don't have. Coach the honest reframe instead.
- **Mock = one question at a time.** Don't dump a list. The value is the pressure and the per-answer critique.

## Related skills
- Reads from: `company-research` (the research file), `story-bank` (stories to map), `job-fit-scoring` (the fit gaps become objection prep), `job-search-setup` (positioning)
- Routes to: `story-bank` when a likely question has no covering story
- Hands off to: `interview-followup` (after the interview: thank-you, debrief, and feed the bank)