---
name: interview-intel
description: >
  Generates a structured interview intel brief for a specific company and role.
  Trigger this skill whenever the user provides a company name, job description,
  and/or resume and wants to prepare for an interview. Covers six outputs:
  (1) Business intelligence on the company — how they make money, bottom line
  metric, key customers, growth trajectory, recent news; (2) Answers to the 5
  key business questions a candidate should know cold; (3) A JD vs resume gap
  analysis — strengths, gaps, and bridge stories; (4) Predicted interview
  questions based on the JD and role level; (5) Smart questions to ask the
  interviewer; (6) A "why this company" answer in the candidate's voice. Use
  this skill any time a user mentions "interview prep", "interview intel",
  "preparing for an interview", "I have an interview at [company]", or shares
  a JD and resume together.
---

# Interview Intel Skill

Generates six outputs for any company + role combination:
1. **Business intelligence brief** — how the company makes money, bottom line metric, key customers, growth trajectory, recent news
2. **Answers to the 5 key business questions** — the questions a candidate must be able to answer to treat the interview as a business partnership
3. **JD vs resume gap analysis** — competency-by-competency match with strengths, gaps, and bridge stories
4. **Predicted interview questions** — what the hiring manager is likely to ask, based on the JD and role level
5. **Questions to ask the interviewer** — tailored to the company's specific moment, not generic
6. **"Why this company" answer** — 3-4 sentences in the candidate's voice, specific to this company's situation

---

## Inputs

Collect these before generating anything. If any are missing, ask for them explicitly:

| Input | Required | How to get it |
|-------|----------|---------------|
| Company name | Yes | Ask user |
| Company website | Preferred | Ask user, or infer from company name |
| Job description | Yes | User pastes or uploads |
| Resume | Yes | Auto-discover (see below), or ask user |

### Resume auto-discovery

Before asking the user for their resume, search the current working directory for it automatically. Check for these filenames in order:

1. `resume.pdf`
2. `resume.docx`
3. `resume.md`
4. `cv.pdf`
5. `cv.docx`

If a match is found, read it and confirm to the user: *"I found resume.pdf in the current directory — using that. Let me know if you'd like to use a different one."* Then proceed without waiting for a response.

If no resume is found in the current directory, ask the user to paste it or provide a file path.

---

## Step 1 — Business Intelligence Brief

### Research process
1. Web search: `[company] ARR revenue growth 2024 2025`
2. Web search: `[company] funding valuation investors`
3. Web search: `[company] key customers enterprise`
4. Web search: `[company] recent news acquisitions product launches`
5. Fetch company website homepage and About/Customers pages if available

### Output format

Write as flowing prose, not bullet points. Sections:

**How [Company] makes money**
Business model in plain terms. Is it B2B SaaS, marketplace, usage-based, services? Who pays and for what?

**The bottom line metric**
What does the company optimize for — ARR, GMV, DAUs, revenue, downloads, impact? Cite the most recent number available. If private with no public data, explain the likely metric based on business model.

**Key customers and verticals**
Named enterprise customers if available. Industry verticals they serve. What problem they solve for each.

**Growth trajectory**
Funding history, ARR growth if available, headcount signals, geographic expansion, recent product launches.

**Recent news worth name-dropping**
1-3 specific recent events (acquisitions, partnerships, product launches, executive hires) with brief strategic interpretation — why does this matter?

**How to frame this in the interview**
One paragraph connecting the business context to the candidate's role specifically. How does the role contribute to the bottom line metric?

---

## Step 2 — The 5 Key Business Questions

Generate a short, speakable answer (2-4 sentences, conversational tone) for each. These should be answers the candidate can say out loud, not read from a slide.

1. **How does the company make money?**
2. **What's their bottom line — revenue, growth, downloads, impact?**
3. **How does the team I'd join contribute to that bottom line?**
4. **How would my specific role contribute?**
5. **Can I walk into an interview and talk about the business, not just the tasks?** *(This one is a self-check — answer it as a brief yes/no + one sentence summary of the business case for the candidate's specific fit)*

Format each as:
```
**Q: [question]**
[2-4 sentence answer in the candidate's voice, using their background]
```

---

## Step 3 — JD vs Resume Gap Analysis

### Process
1. Extract every explicit requirement from the JD — required skills, experience, behaviors, qualifications
2. Extract every signal from the resume — roles, technologies, outcomes, responsibilities
3. Map them against each other competency by competency

### Output format

Render as a table with these columns:

| Competency Area | Required Skills (from JD) | Priority | Candidate Evidence (from resume) | Assessment |
|----------------|--------------------------|----------|----------------------------------|------------|

**Priority levels:**
- `High` — explicitly required, mentioned multiple times, or core to the role title
- `Medium` — mentioned once, or implied by the role
- `Low` — listed as "a plus" or "nice to have"

**Assessment values:**
- `Strong match` — direct evidence in resume, name the specific experience
- `Indirect match` — transferable experience exists, needs a bridge story
- `Gap — address` — no direct evidence, listed as required; provide a one-line bridge
- `Gap — minor` — no evidence but listed as "a plus"; note briefly

### After the table

Write a **3-sentence summary**:
- Sentence 1: Overall fit signal ("Your background maps strongly to X of Y core requirements")
- Sentence 2: The strongest differentiator ("Your most distinctive asset for this role is...")
- Sentence 3: The one thing to prepare ("The only gap worth preparing a bridge story for is...")

---

## Step 4 — Predicted Interview Questions

Based on the JD, role level, and company stage, generate the questions this hiring manager is most likely to ask. Tailor them to the specific context — a Series B engineering hire gets different questions than a recruiter screen at a public company.

Organize into three groups:

**Opener / background**
The questions that almost always come first. Generate 3-4 likely questions and a one-sentence coaching note for each on what the interviewer is really listening for.

**Role-specific**
Questions that probe the core competencies in the JD — technical skills, leadership, cross-functional work. Generate 4-6, drawn directly from the requirements listed. For each, note what a strong answer demonstrates.

**Judgment and tradeoffs**
The harder questions that separate senior candidates from junior ones — how they handle ambiguity, conflict, failure, or competing priorities. Generate 2-3 based on the seniority level implied by the JD.

Format each as:
```
**"[Question]"**
*What they're listening for: [one sentence]*
```

---

## Step 5 — Questions to Ask the Interviewer

Generate 6-8 questions the candidate can ask at the close of the interview. These should be grounded in the specific business intelligence gathered in Step 1 — not generic "what's the culture like" questions.

Split into two groups:

**Business and strategy (3-4 questions)**
Questions that demonstrate the candidate has done real research — referencing the company's funding stage, recent news, product direction, or competitive position. Make them specific enough that a generic candidate couldn't ask them.

**Role and team (3-4 questions)**
Questions about how the team operates, what success looks like in the first 90 days, how the role connects to the company's priorities, and what the biggest challenge is. These signal ownership mindset.

For each question, add a one-line note on why it's a strong question to ask:
```
**"[Question]"**
*Why ask this: [one sentence]*
```

---

## Step 6 — "Why This Company" Answer

Before writing anything, run a short Socratic dialogue to surface the candidate's genuine motivation. Do not skip this — a generated answer without it will sound researched but not real.

### Questioning process

Ask 2-3 questions, **one at a time**. Wait for the candidate's answer before asking the next. Adapt each question based on what they just said — don't follow a rigid script.

Start with a question that opens space for honest motivation, not rehearsed positioning. Good starting questions:
- *"What made you apply to this one specifically — was there a moment where it clicked?"*
- *"What is it about what they're building that actually interests you, beyond the role itself?"*
- *"Is there something in your own experience that makes this problem space feel personal?"*

Then follow up based on their answer:
- If they mention the product or mission → ask what specifically about it resonates and why now
- If they mention their background → ask what they want to do differently or more of in this role
- If their answer is vague → gently push: *"What would make this feel like the right move a year from now?"*

Stop at 3 questions maximum. If a rich, genuine answer has already surfaced after 2, stop there.

### Writing the answer

After the dialogue, write a 3-4 sentence answer in the candidate's voice using what they actually said — not just what the resume and business intel suggest.

Requirements:
- Must use something the candidate said in the dialogue, not just researched facts
- Must reference something specific about the company's current moment — a funding round, a product direction, a market they're going after
- Must sound like a person speaking, not a cover letter
- Should be conversational enough to open a dialogue, not close one

Format as a single paragraph the candidate can read, internalize, and adapt. Follow it with one line noting which part came from their own words: *"This answer uses what you said about [X] — that's the part that makes it yours."*

---

## Step 7 — Save Output to File

After generating all three parts, save the full output to disk:

1. Determine the save location using this order of preference:
   - A path the user explicitly provided (e.g. "save to ~/interviews")
   - The current working directory where the skill was invoked
   - If neither is clear, ask the user: "Where should I save this? (I'll create a `[Company Name]/overview.md` file there)"
2. Create a folder named after the company inside that location (replace spaces with hyphens if needed)
3. Write the complete output — all three parts — into `[save location]/[Company Name]/overview.md`
4. Include a metadata header at the top of the file:

```
# Interview Intel: [Company Name] — [Role Title]
```

5. Confirm to the user that the file has been saved and show the full path.

---

## Formatting rules

- Business intelligence and business questions: flowing prose, conversational, written in second person where helpful ("You want to be able to say...")
- Gap analysis: table format
- No excessive headers or nested bullet points within sections
- Cite specific data points (ARR numbers, customer names, funding rounds) wherever available
- If data is unavailable or the company is private with limited public info, say so explicitly rather than fabricating

---

## Example trigger prompts

- "I have an interview at Stripe next week, here's the JD and my resume"
- "Help me prep for my interview — company is Notion, here's the job description"
- "I'm interviewing at a startup called Retool, can you research them and compare to my background?"
- "Prep me for this role" + attached JD + attached resume