Job Hunt Interview Prep: Install, Source and Security | FunnelSlayer

Job Hunt Interview Prep

Published by ibrahimkobeissy in ai-second-brain-template

No known issues14 installs

What this skill does

Build your general interview-prep doc — walk the top known interview questions (HR/behavioral, technical, management), co-write answers drawn from your story bank, and save to hunter/career/interview-prep.md. Flags questions you have no story for yet (don't fabricate). Use when the user wants to prepare for interviews generally, write/rehearse answers to common questions, or build their Q&A prep.

Add Job Hunt Interview Prep to your agent

Review the source and files first. When you are ready, copy the prompt instruction or use the CLI command supported by your environment.

Install with a prompt

Paste this into a compatible coding agent:

add this skill "job-hunt-interview-prep" from https://github.com/ibrahimkobeissy/ai-second-brain-template

Install with the CLI

Run this command in a controlled environment after reviewing the repository:

npx skills add https://github.com/ibrahimkobeissy/ai-second-brain-template --skill job-hunt-interview-prep

Skill instructions

job-hunt-interview-prep — answers to the questions you'll actually get

Walk the canonical interview questions (HR/behavioral, technical, management) and co-write your answers, drawn from your real story bank — output to hunter/career/interview-prep.md. This is the general prep (not company-specific — that's job-hunt-scout-company). Answers come from hunter/career/my-stories.md; where there's no story for a question, flag the gap instead of inventing one.

Hard rules

  • Answers trace to real stories. Pull from my-stories.md; never fabricate an achievement to answer a question.
  • Flag gaps, don't fill them. A question with no backing story → mark it and suggest running job-hunt-capture-story.
  • One question at a time when co-writing.
  • Private output (hunter/career/, tracked to the private remote, excluded from JH-06 public sync). Don't stage or commit.

Steps

  1. Read context: hunter/personal-profile/cv.md + hunter/career/my-stories.md.
  2. Adapt the bank: use the canonical question set in template.md; tailor the technical block to the domain in cv.md (don't ask backend questions of a designer).
  3. Per question: find the best-fit story → draft a concise answer co-written with the user (propose, ask, refine). For behavioral, keep STAR shape. Record which story each answer draws on.
  4. Gaps: list questions with no backing story; suggest capturing them via job-hunt-capture-story.
  5. Write hunter/career/interview-prep.md from template.md: each Q → your answer → source story (→ STAR for behavioral). Re-running updates it, doesn't duplicate.

Done when

hunter/career/interview-prep.md has a real, story-backed answer per answered question, gaps flagged honestly, nothing invented.

Files included

  • SKILL.md
  • template.md

More skills