| name | interview-prep-builder |
| description | Prepare for upcoming interviews with role-specific talking points. Use this skill when the user says "I have an interview", "prep me for my interview", "interview tomorrow", "interview on Friday", "help me prepare for [company]", "what should I expect?", "what questions will they ask?", "practice interview", or any request to get ready for a specific upcoming interview. Also trigger when the user mentions a scheduled interview and seems nervous or wants to review key points. This skill generates a custom prep document based on the specific role, the user's match score, past interview patterns, and known gaps. Do NOT trigger for post-interview reflection (interview-debrief) or general job search strategy (search-optimizer).
|
| recommended_model | {"default":"sonnet","reasoning":"Generating structured prep materials from metadata, job descriptions, and past interview patterns is well-suited for Sonnet. The questions and talking points follow predictable patterns enhanced by specific data.\n","upgrade_to_opus_when":"The user wants to practice with mock interview questions and needs nuanced follow-up probing, or wants help with difficult gap-bridging narratives.\n"} |
Interview Prep Builder — Walk In Ready
Intent
- Talking points must trace to real achievements — every story, number, and claim must exist in achievements.md or base-resume.docx; preparing the candidate to say something they cannot substantiate is sabotage
- Gap bridges must be genuine — "my AWS work translates to GCP" is honest positioning; "I've been learning GCP on my own" without evidence is fabrication
- Customized per company and role — generic interview prep is available everywhere; the value is connecting THIS candidate's specific achievements to THIS role's specific requirements
- Practice over perfection — the prep doc is a study guide, not a script; the candidate should internalize talking points, not memorize paragraphs
- Confidence from preparation — knowing your best stories and how they connect to the role reduces anxiety and improves performance
- Learning loop from past interviews — previous debrief patterns inform question prediction; prep improves with every interview in the system
- Prep document in 15-20 minutes; 3 talking points minimum; 10-15 likely questions — comprehensive enough to be useful, concise enough to actually be reviewed before the interview
- Past, Present, Future is the default opener — every prep starts with a customized 60-90 second intro using the PPF framework (see
references/past-present-future.md); this format has proven effective across interviews and should always be generated first
Why this skill exists
Going into an interview unprepared means relying on improvisation for the most
important conversations in your job search. This skill does the homework: which
achievements to highlight, how to bridge your gaps, what questions they'll likely
ask based on the role type, and what to ask them.
Prerequisites
- Application folder with
metadata.json and job-description.md
master/achievements.md for pulling specific accomplishments
- Match score must exist — if
metadata.json has no match_score (or it's empty),
prompt the user to run scoring first via job-intake or the web app Score button.
Interview prep without a match score means no gap bridges, no keyword targeting,
and weaker talking points. Don't proceed without it.
- Optional but valuable: previous
interview-notes.md or interview-debrief-rN.md
files from other applications (for pattern-based question prediction)
- Optional:
master/interview-learnings.md for proven tactical lessons from past
interviews
Workflow
Step 1: Identify the interview
Which company, role, and when? Check context. If the user just got scheduled,
they'll mention it. Confirm the details and check that the application folder
exists.
Step 2: Load context
Read:
metadata.json — match score, gaps, keywords, tailoring intensity, interviews[] for prior round context
job-description.md — the full role requirements
company-brief.md — company research (if it exists, from company-research skill)
study-sheet.md — quick-reference study data (if it exists)
master/achievements.md — your accomplishment inventory
master/narrative.md — your positioning themes for this role type
master/story-bank.md — signature stories with pillar tags and deployment history
master/analogy-arsenal.md — memorable metaphors with usage tracking
master/interview-learnings.md — proven tactical lessons from past interviews (if it exists)
- Previous
interview-debrief-rN.md files in this application folder — for THIS company's prior rounds
- Previous
interview-notes.md files across applications (if any) — for cross-company question patterns
Story reuse check: Before selecting stories and analogies for this prep, check
the "Deployed in" fields in master/story-bank.md and master/analogy-arsenal.md.
If prepping for a later round at the same company, do NOT reuse stories or lines
already deployed in earlier rounds. Flag any conflicts and suggest alternatives.
If company-brief.md does not exist, suggest running company-research first for
richer, company-specific talking points and questions.
Match score gate: If match_score.overall is empty/missing, stop and tell the
user: "This application hasn't been scored yet. Run scoring first so I can generate
gap bridges and keyword-targeted talking points. Use the Score button in the web app
or run the job-intake skill."
Step 3: Generate the prep document
Create interview-prep.md in the application folder:
# Interview Prep — [Company] [Role]
**Interview date:** [date]
**Interview type:** [phone/video/in-person/panel]
**Interviewer(s):** [if known]
## Your Opening — Past, Present, Future
> Your 60-90 second intro when they say "Tell me about yourself."
> Practice this out loud until it's natural, not memorized.
**Past:** "[Career arc + biggest proof point with a number. Pull from achievements.md,
emphasizing the experience most relevant to this role.]"
**Present:** "[What you're actively doing now that shows growth and relevance.
Pick the angle that resonates most with this specific role.]"
**Future:** "[Why THIS company specifically. Reference their mission, challenge,
or approach by name. Connect it to your experience. Never generic.]"
> See `references/past-present-future.md` for the full framework and proven example.
## Your Top 3 Talking Points
Based on this role's requirements and your strongest matches:
### 1. [Strength area]
**The story:** [Specific achievement from achievements.md]
**Numbers:** [Quantified impact]
**Connection to role:** [How this maps to their requirement]
### 2. [Strength area]
**The story:** [Specific achievement]
**Numbers:** [Quantified impact]
**Connection to role:** [How this maps]
### 3. [Strength area]
**The story:** [Specific achievement]
**Numbers:** [Quantified impact]
**Connection to role:** [How this maps]
## Gap Bridges
For each addressable gap from the match score:
### [Gap: specific skill/experience]
**Bridge approach:** [How to position adjacent experience]
**Example:** "While I haven't worked specifically with [X], my work on [Y]
at [Company] involved similar [principles/challenges/patterns]..."
**Evidence:** [Specific accomplishment that partially addresses the gap]
## Likely Questions
Based on the role type and past interview patterns:
### Behavioral
- "Tell me about a time you scaled a team from scratch."
→ Use: [specific MedQuest or Red Spot story]
- "Describe a technical decision you made that had significant business impact."
→ Use: [specific achievement]
### Technical
- [Role-specific technical questions based on job description]
### Strategic
- "How would you approach [key challenge from job description]?"
→ Framework: [suggested approach]
### Curveball (from past interviews)
- [Questions that came up in similar role interviews]
## Questions to Ask Them
Show you've done homework and are evaluating them too:
1. "What's the biggest technical challenge the team is facing right now?"
2. "How does this role interact with [specific team/function from job description]?"
3. "What does success look like in the first 90 days?"
4. "What's the team's approach to [specific methodology from job description]?"
## Numbers Cheat Sheet
Quick-reference for your key metrics (from achievements.md):
| Metric | Number | Context |
|--------|--------|---------|
| [Achievement 1] | [Number] | [Company, Year] |
| [Achievement 2] | [Number] | [Company, Year] |
| [Achievement 3] | [Number] | [Company, Year] |
## Lessons from Past Interviews
> Populated from `master/interview-learnings.md` — only include lessons tagged as
> "proven" or directly relevant to this interview type.
- [Relevant proven lesson 1]
- [Relevant proven lesson 2]
- [Relevant "what to improve" item if applicable]
## System Design Scenarios
> Include this section ONLY when interview type is `system_design` or `technical_panel`.
> Generate 2-3 design scenarios relevant to the company's domain.
### Scenario 1: [Domain-relevant problem]
**Problem:** [1-2 sentence description]
**Components:** [Key building blocks — API gateway, queue, DB, cache, etc.]
**Key trade-offs:** [2-3 decision points with pros/cons]
**Your experience connection:** [How your past work relates]
### Scenario 2: [Another domain-relevant problem]
...
## Pre-Interview Checklist
- [ ] Practice your Past, Present, Future opener out loud (aim for 60-90 seconds)
- [ ] Review the job description one more time
- [ ] Review company-brief.md and study-sheet.md (run company-research if missing)
- [ ] Confirm interview logistics (link, time zone, who you're meeting)
- [ ] Have your talking points fresh in mind
- [ ] Prepare 3-4 questions for them
- [ ] Have a copy of your tailored resume handy
- [ ] **If system design/technical panel:** Drawing tool ready (Excalidraw, Miro, etc.)
- [ ] **If system design/technical panel:** Practice sketching one scenario before the call
Step 4: Customize for interview round
Round escalation strategy: Each round serves a different purpose. The intensity
and depth of your content should escalate naturally:
- R1 = rapport + high-level fit (they decide if you're worth more time)
- R2 = depth + proof (they validate your claims with stories and numbers)
- R3+ = technical calibration + vision (they test your ceiling)
Recruiter screen (Round 1):
- Deliver the full PPF opener — this is where it shines most
- Focus on: fit, salary alignment, high-level background
- Keep stories short (1-2 minutes each), use short versions from
master/story-bank.md
- Emphasize enthusiasm for the role
- Recruiter-as-ally: Invest 60 seconds in a genuine personal connection — find a
bonding moment (shared interest, side project, anything human). Recruiters who like
you become advocates: they feed you intel about the hiring manager's priorities,
timeline pressure, and what to emphasize. This paid off at Ensemble — Bethany
connected over F1 and Easter Island, then provided detailed insider context for R2.
Hiring manager (Round 2):
- Deliver the full PPF opener, tuned toward leadership and strategy
- Focus on: leadership, strategic thinking, team building
- Share detailed stories with numbers — use full versions from
master/story-bank.md
- Select analogies from
master/analogy-arsenal.md that match the discussion topic
- Ask about their vision and challenges
Technical panel (Round 3+):
- Shorten PPF to 30-45 seconds — they want to get to technical depth quickly
- Focus on: architecture decisions, technical trade-offs
- Be ready for whiteboard or system design
- Show depth in their specific tech stack
- Use the "Define the Term First" technique for strategic questions (see
references/common-questions.md)
Final round / executive:
- Deliver the full PPF opener, emphasizing business impact in Past and mission alignment in Future
- Focus on: business impact, cultural fit, long-term vision
- Ask about company strategy and growth
- Show alignment with company mission
Step 5: Incorporate past interview learning
From this company's prior rounds (check interviews[] in metadata and
interview-debrief-rN.md files in the application folder):
- What topics were already covered? Don't repeat the same stories.
- What did the interviewer probe on? Expect follow-up depth.
- What intel was gathered? Use it in your questions and positioning.
From cross-company patterns (check interview-notes.md across other apps):
- Pull recurring questions from similar roles
- Note which stories/approaches worked well
- Highlight gap bridges that succeeded or failed
From master/interview-learnings.md (the accumulated knowledge base):
- Include "proven" lessons (3+ occurrences) in the "Lessons from Past Interviews" section
- Include "what to improve" items relevant to this interview type as reminders
- Match lessons by interview type: system_design, behavioral, recruiter_screen, etc.
System design detection: If the interview type (from interviews[] or user input)
is system_design or technical_panel, generate the "System Design Scenarios" section:
- 2-3 scenarios relevant to the company's domain and the role's focus area
- Include component diagrams (ASCII or description), key trade-offs, and connection to candidate's experience
- Reference the company's tech stack from company-brief.md or job-description.md
Step 6: Present the prep
Share the document. Ask if there are specific areas the user wants to drill
deeper on. Offer to do mock Q&A for the gap bridge sections.
PPF practice: Suggest the candidate read their Past, Present, Future opener
out loud 2-3 times. It should feel conversational, not rehearsed. Time it —
if it's over 90 seconds, trim.
Edge cases
- No interview date set — Create the prep anyway. Better to be ready early.
- No past interviews — Generate questions from the job description and role
type alone. Flag that predictions will improve after debriefs.
- Second interview at same company — Reference the first interview debrief.
Focus on what they're likely to probe deeper.
- Panel with unknown members — Prepare for multiple perspectives: technical,
managerial, product. Cover all angles.
Skill composition
| Upstream | This skill | Downstream |
|---|
| job-intake (match score, gaps) | interview-prep-builder (preparation) | interview-debrief (captures what happened) |
| company-research (company intel) | | |
| interview-debrief (past patterns) | | |
Read references/common-questions.md for question frameworks, answer techniques, and role-type patterns.
Read references/past-present-future.md for the PPF opener framework and proven example.
Read master/story-bank.md for signature stories tagged by pillar with deployment tracking.
Read master/analogy-arsenal.md for memorable metaphors with usage rules.