interview-prep

name	interview-prep
description	Maps your stories to interview rounds based on JD requirements. Generates per-round talking points, mock questions with scoring, and coaching notes based on personality profile. Say "prep me for [Company]" to generate a prep doc, or "mock interview for [Company]" to practice with feedback.
triggers	["prep me for","ip","interview prep","mock interview","practice for","get ready for"]

Interview Prep — Career OS Skill

Purpose

Prepares you for interviews by matching your stories to the role's requirements, generating prep documents with talking points, and running mock practice sessions with structured feedback.

Output Format

Always start with:

━━━ Career OS: Interview Prep ━━━

How to Invoke

• prep me for [Company] — generate a prep document
• mock interview for [Company] — practice with scored feedback
• interview prep — show what preps are available

---

DATA ARCHITECTURE

Inputs

Source	Path	What It Provides
Target company/role	User request or pipeline entry	Which interview to prepare for
Stories	brain/stories/*.md	Career stories with competency frontmatter
Story Index	brain/stories/STORY_INDEX.md	Competency clusters for fast matching
Pipeline	career-intelligence/projects/job-search/job-pipeline.json	Stage, contacts, next steps for target
JD	brain/reference/jd-samples/{company}*.pdf	Role requirements, competencies sought
Personality	brain/config/personality.md (optional)	RHETI type for coaching calibration
Skills matrix	brain/identity/skills-matrix.md	Technology proficiency for technical rounds
JD Alignment Framework	brain/projects/jd-alignment-framework.md	Track definitions and requirements — used for story-to-track mapping and round preparation

Brain API (brain-kernel >= 1.0.0)

All writes go through brain.write(). interview-prep/ maps to the engine's projects/** owned namespace. Mock debriefs land in stories/**.

Outputs

Output	brain.write() path	When Created
Prep document	career-intelligence/projects/interview-prep/prep-{company}.md	Every prep session
Mock debrief	career-intelligence/stories/mock-{company}-{date}.md	After mock session

Legacy Read Tolerance

Canonical write path: career-intelligence/projects/interview-prep/prep-{company}.md.

The skill may encounter files that pre-date this convention or were written outside the skill (user-initiated notes, legacy installs). On read:

• Accept any *.md file in brain/interview-prep/ for the requested company — match by filename substring (lowercased).
• Prefer prep-*.md when multiple matches exist; fall back to any .md match if no canonical file is present.
• Ignore files in _archive/ unless explicitly asked ("show archived prep for X").
• intel-*.md files are insider-intel notes (role context, interviewer background) — read alongside prep docs when available; do not generate intel-*.md files from prep sessions (that's a separate input class).

Do NOT silently rewrite legacy filenames on read. Filename migration is handled by plugin migrations (migrations/vN-to-vN+1.sh), not by the skill at runtime. This keeps the skill's write path canonical and the migration the single source of truth for filename shape changes.

---

BEHAVIOR: Prep Mode (prep me for [Company])

Step 1: Gather Context

1. Read pipeline entry for the company — get role, stage, interviewer names, dates
2. Read stored JD from brain/reference/jd-samples/ — extract key requirements grouped by category (technical, leadership, domain, culture)
3. Read STORY_INDEX.md — match competency clusters to JD categories
4. Read matched story files for metrics and talking points

Step 2: Generate Prep Doc

Write via brain.write("career-intelligence/projects/interview-prep/prep-{company}.md", ...):

# Interview Prep — {Company} ({Role})
**Stage:** {current stage from pipeline}
**Interviewer:** {name if known}
**Date:** {if known from pipeline}
**Generated:** {today's date}

## Role Summary

{2-3 sentences: what they're looking for, mapped to your profile}

### Strengths (direct matches)
- {requirement} → {your evidence with metric}

### Gaps (no direct story)
- {requirement} → Suggested angle: {framing advice}

## Story Map

| JD Requirement | Your Story | Key Talking Points |
|---|---|---|
| {requirement 1} | {story title} | • {metric/outcome} • {action taken} • {scale/scope} |
| {requirement 2} | {story title} | • {metric/outcome} • {action taken} |
| {requirement 3} | ⚠️ No matching story | Angle: {suggested framing using adjacent experience} |

## Round-by-Round Guide

### Behavioral Round
- Lead with: {strongest story for this company}
- Watch for: {common behavioral questions for this role level}
- Your edge: {what makes your background unique for this role}

### Technical/System Design Round
- Technologies to highlight: {from skills-matrix, matched to JD}
- Architecture story: {best system design story}
- Gaps to prepare for: {technologies in JD not in your matrix}

### Leadership Round
- Team scale story: {largest team managed, with metrics}
- Cross-functional story: {best cross-org collaboration}
- Failure/recovery story: {shows resilience and learning}

## Questions to Ask

Tailored to role level, company stage, and interviewer background:
- {question} — *Why: {what this reveals about the role/company}*
- {question} — *Why: {shows you've done research}*
- {question} — *Why: {evaluates culture fit for YOU}*

## Coaching Notes

{Generated from personality profile if available, otherwise from observed patterns}
- {tip based on RHETI type or common patterns for this user}
- {interviewer-specific adjustment if interviewer background known}

Step 3: Present Summary

After writing the prep doc, present a condensed version in conversation:

━━━ Career OS: Interview Prep ━━━

Prep doc ready for {Company} — {Role}

STORY MAP: {N} requirements matched, {N} gaps identified
STRONGEST MATCH: {story} → {requirement}
BIGGEST GAP: {requirement} — suggested angle: {1-line framing}

Full prep: career-intelligence/projects/interview-prep/prep-{company}.md

Want to run a mock round? Say "mock interview for {Company}"

---

BEHAVIOR: Mock Mode (mock interview for [Company])

Step 1: Setup

1. Load prep doc (generate via Prep Mode if missing)
2. Ask: "Which round? behavioral / technical / system design / leadership"

Step 2: Conduct Mock

3. Ask questions one at a time — wait for user's full answer before proceeding
4. Questions should be realistic for the role level and company
5. Draw from JD requirements and gap analysis in prep doc

Step 3: Score Each Answer

After each answer, provide structured feedback:

━━━ Feedback ━━━

STAR Structure:  ✅ Situation  ✅ Task  ✅ Action  ⚠️ Result (add the metric)
Specificity:     🟡 — "improved performance" → say "reduced p99 latency from 800ms to 120ms"
Relevance:       ✅ — directly answers the question
Conciseness:     ✅ — good length for a 2-minute answer

💡 You mentioned the outcome but not the scale — add the 40% metric
   and the team size (12 engineers across 3 teams).

Ready for the next question? (q to stop)

Scoring dimensions:

• STAR structure — did they hit Situation, Task, Action, Result?
• Specificity — concrete metrics, names, numbers vs vague claims
• Relevance — does the answer actually address what was asked?
• Conciseness — under ~2 minutes of speaking time?

Step 4: Debrief

After 3-5 questions (or when user says q/done):

━━━ Mock Debrief ━━━

Questions: {N} asked
Average score: {rating}
Strongest answer: {question topic} — {why it worked}
Needs work: {question topic} — {specific improvement}

Debrief saved to brain/stories/mock-{company}-{date}.md

Write debrief via brain.write("career-intelligence/stories/mock-{company}-{date}.md", ...) with frontmatter:

---
type: mock-interview
company: {company}
date: {date}
round: {round type}
questions_asked: {N}
---

---

BEHAVIOR: No Stories Available

If brain/stories/ is empty or has no STORY_INDEX.md:

━━━ Career OS: Interview Prep ━━━

You don't have stories indexed yet. Stories power interview prep —
they're how I match your experience to what the company is looking for.

Say "organize" to index your career stories first, then come back.

---

BEHAVIOR: No JD Available

If no JD in brain/reference/jd-samples/ for this company:

• Generate prep based on pipeline entry + role title + typical requirements for that level
• Note in prep doc header:

  ⚠️ No JD available — prep based on role-level norms.
  Paste the JD for a targeted version: "here's the JD for {Company}"
  

• Mock questions will be more generic but still calibrated to role level

---

BEHAVIOR: Re-run Prep

If a prep doc already exists for this company:

• Overwrite with fresh version (previous version in git history)
• Note: "Updated prep for {Company} — previous version in git history"

---

Dependencies

• organize — stories need frontmatter for competency matching (required)
• Pipeline entry — helpful but not required (can work from JD alone)
• job-search-scheduler — provides stored JDs (helpful, not required)