Agent skill

interview-feedback

Post-PM-interview debrief and continuous improvement for job search

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/interview-feedback

SKILL.md

Note: This skill is for PM career interviews (job interviews). For debriefing after conducting user research interviews, use /user-research-synthesis instead.

Interview Feedback Skill

Capture insights immediately after PM job interviews. Quick retrospective to improve for next rounds and future interviews.

Quick Start

/interview-feedback

Just finished a PM interview? Let's debrief while it's fresh.

Tell me:
1. Which company and role? (e.g., "Senior PM at Stripe")
2. What questions did they ask? (as many as you remember)
3. How do you think you did? (gut feeling: great / good / okay / rough)

I'll help you:
- Score your performance across 5 key dimensions
- Identify specific strengths to repeat
- Create an improvement plan for weak areas
- Track patterns across multiple interviews

For user research interview debriefs, use /user-research-synthesis instead.

Output: outputs/interview-feedback/[date]-[company]-debrief.md

When to Use This Skill

  • Within 15 minutes of finishing a PM job interview
  • After phone screens, onsite rounds, or take-home presentations
  • When you advance to next round (capture what worked)
  • When you get rejected (learn for next time)

Workflow

Step 0: Emotional Check-In

Before scoring, acknowledge the emotional state. Interviews are stressful.

Ask: "Before we get into the analysis -- how are you feeling about it? Quick gut reaction."

Why this matters:

  • PMs who feel terrible often performed better than they think (impostor syndrome)
  • PMs who feel great sometimes missed signals (overconfidence)
  • Naming the emotion first prevents it from biasing the objective scoring

Use their response to calibrate:

  • If they say "awful" -- be extra careful to highlight what went well before diving into gaps
  • If they say "great" -- still praise strengths but don't skip honest critique
  • If they say "mixed" -- mirror that by leading with the strongest and weakest dimensions

Step 1: Immediate Brain Dump

Right after the interview ends, capture everything while fresh:

## Raw Notes (Stream of consciousness)

**Questions I was asked:**
1. [Question 1 - exact wording if you remember]
2. [Question 2]
3. [Question 3]
[List all questions you can recall]

**My answers (rough outline):**
- Q1: [What I said - framework used, key points made]
- Q2: [What I said]
- Q3: [What I said]

**Follow-up questions interviewer asked:**
- [Follow-up 1 - signals they wanted more depth here]
- [Follow-up 2]

**Interviewer reactions I noticed:**
- [When they nodded, took notes, seemed engaged]
- [When they seemed confused or skeptical]
- [When they cut me off or moved on quickly]

**Things that felt good:**
- [Moment 1: e.g., "Nailed the metrics definition"]
- [Moment 2]

**Things that felt rough:**
- [Moment 1: e.g., "Stumbled on prioritization rationale"]
- [Moment 2]

Step 2: Structured Self-Assessment (5 Key Dimensions)

Rate yourself 1-5 on these five core dimensions. Be honest -- this is for your growth, not your ego.

## Performance Scorecard

**Interview Type:** [Product Sense / Execution / Behavioral / Design / Technical]

**Company:** [Company name]
**Role:** [Role title]
**Interviewer:** [Name + title if you know it]
**Duration:** [Scheduled time] → [Actual time]

### The 5 Core Dimensions (1-5 scale)

**1. Framework Usage:** ___/5
- Did I use a clear structure? (5-Step, CIRCLES, AARM, STAR)
- Was my thinking organized and easy to follow?
- Did I signal my structure upfront? ("I'll approach this in 3 parts...")
- 1 = No structure, stream of consciousness
- 3 = Some structure but lost it halfway
- 5 = Crystal clear framework, interviewer could follow without notes

**2. Specificity:** ___/5
- Did I use real examples, data, and concrete details?
- Were my metrics specific? ("3.2% conversion" not "low conversion")
- Did I name real products, companies, or technologies?
- 1 = Entirely hypothetical, vague generalities
- 3 = Some specific examples but padded with hand-waving
- 5 = Every claim backed by real data or concrete examples

**3. Creativity:** ___/5
- Did my answer stand out from what 100 other candidates would say?
- Did I show unique insight or a surprising angle?
- Did I go beyond the obvious first solution?
- 1 = Textbook answer anyone could give
- 3 = Solid but predictable
- 5 = Genuinely surprising insight that made the interviewer lean in

**4. Communication:** ___/5
- Was I clear and concise?
- Did I avoid rambling, filler words, or hedging?
- Did I read the room and adjust?
- 1 = Rambled, lost the thread, interviewer looked confused
- 3 = Generally clear but some tangents
- 5 = Every sentence earned its place, confident delivery

**5. Product Sense:** ___/5
- Did I demonstrate user empathy?
- Did I connect user value to business value?
- Did I show understanding of trade-offs?
- 1 = Talked about features without understanding users
- 3 = Mentioned users but didn't deeply empathize
- 5 = Interviewer said "great point" or visibly reacted to insight

**Overall Performance:** [Average of 5 scores] / 5

How to interpret your scores:

Average What it means What to do
4.5-5.0 Outstanding -- you're likely advancing Maintain and fine-tune
3.5-4.4 Solid -- competitive but not guaranteed Focus on your lowest dimension
2.5-3.4 Needs work -- probably won't advance Intensive practice on 2+ dimensions
Below 2.5 Major gaps -- step back and prepare more Consider delaying next interview by 1-2 weeks

Step 3: What Went Well (2 minutes)

Capture your wins to repeat them:

## Strengths This Interview

✅ **Strength 1:** [What you did well]
   - Example: "Used 5-step Product Sense framework perfectly"
   - Why it worked: "Gave clear structure, interviewer followed easily"
   - Repeat next time: Yes

✅ **Strength 2:** [What you did well]
   - Example: "Asked great clarifying questions upfront"
   - Why it worked: "Avoided solving wrong problem"
   - Repeat next time: Yes

✅ **Strength 3:** [What you did well]
   - Example: "Specific metrics with targets (DAU/MAU 40% → 50%)"
   - Why it worked: "Showed data-driven thinking"
   - Repeat next time: Yes

Step 4: What Needs Improvement (2 minutes)

Be brutally honest about weaknesses:

## Areas for Improvement

❌ **Weakness 1:** [What didn't go well]
   - Example: "Spent too long on problem definition (10 min)"
   - Impact: "Only 5 min left for solutions - felt rushed"
   - Fix for next time: "Cap problem def at 5 min, use timer"

❌ **Weakness 2:** [What didn't go well]
   - Example: "Didn't tie solution back to business metrics"
   - Impact: "Answer felt incomplete, interviewer asked about ROI"
   - Fix for next time: "Always end with user value → business value connection"

❌ **Weakness 3:** [What didn't go well]
   - Example: "Used vague language ('users might like this')"
   - Impact: "Sounded uncertain, not data-driven"
   - Fix for next time: "Replace 'might' with 'based on [data], I expect [outcome]'"

Step 5: Interviewer Signals (1 minute)

Read between the lines:

## Interviewer Engagement Signals

**Positive signals I noticed:**
- ✅ [Signal 1: e.g., "Took detailed notes during my metrics section"]
- ✅ [Signal 2: e.g., "Asked follow-up questions showing genuine interest"]
- ✅ [Signal 3: e.g., "Smiled and nodded when I explained prioritization"]

**Concerning signals I noticed:**
- ⚠️ [Signal 1: e.g., "Cut me off twice - may have been rambling"]
- ⚠️ [Signal 2: e.g., "Didn't ask follow-ups on my solution"]
- ⚠️ [Signal 3: e.g., "Checked phone during my answer"]

**Questions they asked at the end:**
- [Question 1: e.g., "Tell me about a time you failed"]
  → Meaning: [They want to test humility/learning]
- [Question 2]

**My questions I asked:**
- [My question 1]
  → Their response: [Summary]
- [My question 2]
  → Their response: [Summary]

Step 5.5: Prep vs Performance Comparison

If the PM ran /interview-prep before this interview, compare:

Planned strengths vs Actual:

  • What frameworks did you plan to use? Did you use them?
  • What company-specific research did you prepare? Did it come up?
  • What stories did you plan to tell? Did you tell them?

Format in output:

Planned (from /interview-prep) Actual Performance Gap
Use CIRCLES for product sense Used CIRCLES clearly None
Cite company's 3M DAU metric Forgot to mention specific numbers Specificity gap
Tell DataStack churn reduction story Told it but rambled (3 min vs planned 2 min) Communication gap

This comparison accelerates learning by making the gap between intention and execution visible.


Step 6: Pattern Recognition (Multi-Interview)

Before generating feedback, check outputs/interview-feedback/ for previous debrief files.

If previous debriefs are found, show a trend analysis:

## Cross-Interview Pattern Analysis

**Previous debriefs found:** [N] files in outputs/interview-feedback/

### Trend Report (across last [N] interviews)

| Dimension | Interview 1 | Interview 2 | Interview 3 | Trend |
|-----------|-------------|-------------|-------------|-------|
| Framework Usage | X/5 | X/5 | X/5 | Improving / Flat / Declining |
| Specificity | X/5 | X/5 | X/5 | Improving / Flat / Declining |
| Creativity | X/5 | X/5 | X/5 | Improving / Flat / Declining |
| Communication | X/5 | X/5 | X/5 | Improving / Flat / Declining |
| Product Sense | X/5 | X/5 | X/5 | Improving / Flat / Declining |

**Consistent strengths across interviews:**
- [Pattern 1: e.g., "Always strong on user segmentation"]
- [Pattern 2: e.g., "Metrics definition consistently praised"]

**Recurring weaknesses (these need focused practice):**
- [Pattern 1: e.g., "Keep running over time on execution questions"]
- [Pattern 2: e.g., "Weak on competitive analysis section"]

**Company-specific insights:**
- [Company A]: [What they cared most about]
- [Company B]: [What they cared most about]

**Interview type insights:**
- **Product Sense:** [What works for me, what doesn't]
- **Execution:** [What works for me, what doesn't]
- **Behavioral:** [What works for me, what doesn't]

If no previous debriefs exist, note: "This is your first tracked debrief. After 3+ debriefs, I'll show you trend analysis and patterns."


Step 7: What to Improve for Next Time (Specific Drills)

Based on your lowest-scoring dimensions, here are targeted practice drills:

## Improvement Drills by Dimension

### If low on Framework Usage (scored 1-3):
**Practice drill:** Pick 3 products you use daily. For each, answer "How would you
improve this product?" using the 5-Step framework. Time yourself at 25 min each.
- Day 1: Practice with timer, write out structure before speaking
- Day 2: Practice out loud without notes, just framework cards
- Day 3: Record yourself answering, review for structure breaks
**Key habit:** Always say your framework out loud before diving in:
"I'll approach this in 4 steps: first the user, then the problem, then solutions, then metrics."

### If low on Specificity (scored 1-3):
**Practice drill:** Prepare 5 STAR stories with exact metrics.
- For each story, fill in: "This resulted in [X]% improvement in [metric],
  affecting [N] users, saving/generating $[amount]"
- If you don't have exact numbers, estimate and say "approximately"
- Research 5 products: know their DAU, revenue, growth rate, key metrics
**Key habit:** Before every claim, ask yourself: "Can I put a number on this?"

### If low on Creativity (scored 1-3):
**Practice drill:** For any product question, force yourself to generate 5 solutions
before evaluating. At least 2 must be "non-obvious."
- Study product teardowns (Lenny's Newsletter, First Round Review)
- Practice "second-order thinking": "If we do X, then Y happens, which causes Z"
- Look at how other industries solved similar problems
**Key habit:** After your first solution idea, ask: "What would surprise the interviewer?"

### If low on Communication (scored 1-3):
**Practice drill:** Record yourself answering 3 questions. Review for:
- Filler words count (um, uh, like, you know)
- Longest sentence (aim for under 20 words)
- Did you pause before answering? (good) or start rambling? (bad)
- Eliminate hedging: replace "I think maybe" with "I believe" or "The data suggests"
**Key habit:** Take a 5-second pause before answering. Silence is better than filler.

### If low on Product Sense (scored 1-3):
**Practice drill:** For 5 products, write a one-paragraph user story:
"[User type] needs to [job] because [motivation]. Today they [current solution]
which is painful because [friction]. A great solution would [outcome]."
- Read 10 user reviews on App Store/G2 for products you're interviewing about
- Practice explaining WHY a feature exists, not just WHAT it does
**Key habit:** Start every answer with the user, not the product.

Step 8: Action Items for Next Interview (1 minute)

Turn insights into concrete improvements:

## Next Interview Action Items

**Before next interview:**
- [ ] [Action 1: e.g., "Practice 3 execution questions under 25 min"]
- [ ] [Action 2: e.g., "Prepare 2 data points about competitor products"]
- [ ] [Action 3: e.g., "Record myself answering and check for filler words"]

**During next interview:**
- [ ] [Behavior 1: e.g., "Use timer to cap problem def at 5 min"]
- [ ] [Behavior 2: e.g., "Always ask 'How much time do I have?' upfront"]
- [ ] [Behavior 3: e.g., "End every answer with business impact"]

**Follow-up needed:**
- [ ] [Task 1: e.g., "Send thank you email within 24 hours"]
- [ ] [Task 2: e.g., "Connect with interviewer on LinkedIn"]
- [ ] [Task 3: e.g., "Research their recent product launches for next round"]

Output Format

markdown
# Interview Debrief: [Company] - [Interview Type]

**Date:** [Date] | **Duration:** [X min] | **Interviewer:** [Name/Role]

---

## Quick Summary

**Overall feeling:** [Great / Good / Okay / Rough]

**Overall score:** ⭐⭐⭐⭐ (4/5)

**Advance to next round?** [Likely / Maybe / Unlikely]

---

## Questions Asked

1. **[Question 1]**
   - My approach: [Framework used, structure]
   - Score: ⭐⭐⭐⭐⭐
   - What went well: [Specific strength]
   - What to improve: [Specific weakness]

2. **[Question 2]**
   - My approach: [Framework used, structure]
   - Score: ⭐⭐⭐
   - What went well: [Specific strength]
   - What to improve: [Specific weakness]

[Continue for all questions]

---

## Performance Breakdown

| Dimension | Score | Notes |
|-----------|-------|-------|
| Structure & Clarity | ⭐⭐⭐⭐ | Used framework but could be crisper |
| Depth & Rigor | ⭐⭐⭐⭐⭐ | Great data points, specific examples |
| Time Management | ⭐⭐⭐ | Went over on Q1, rushed on Q2 |
| Metrics & Data | ⭐⭐⭐⭐⭐ | Specific targets, good rationale |
| Communication | ⭐⭐⭐⭐ | Clear but used filler words |

---

## What Worked

✅ Used 5-step Product Sense framework consistently
✅ Asked clarifying questions before diving in
✅ Specific metrics with targets (not vague)

---

## What to Improve

❌ Time management - spent too long on problem definition
❌ Didn't connect solution to business metrics clearly
❌ Used tentative language ("might", "could") instead of confident

---

## Interviewer Signals

**Positive:**
- Took detailed notes during metrics section
- Asked thoughtful follow-ups

**Concerning:**
- Seemed to check phone during answer
- Cut me off once (may have been rambling)

---

## Action Items

**For next interview:**
- [ ] Practice with timer - cap sections at target time
- [ ] Always end with "user value → business value" connection
- [ ] Replace tentative language with data-backed confidence

**Follow-up:**
- [ ] Send thank you email by [date]
- [ ] Connect on LinkedIn

---

**Next interview:** [Date if scheduled, or "TBD"]
**Next interview type:** [Product Sense / Execution / etc.]

Special Cases

After Rejection

## Rejection Debrief

**Rejection type:** [After screening / After onsite / After final round]

**Official reason given:** [If any]

**My hypothesis on why:**
- [Hypothesis 1: e.g., "Weak on technical depth questions"]
- [Hypothesis 2: e.g., "Better candidate with more relevant experience"]
- [Hypothesis 3: e.g., "Company culture fit concerns"]

**What I learned:**
- [Learning 1]
- [Learning 2]

**How I'll improve:**
- [Specific action 1]
- [Specific action 2]

**Apply these learnings at:** [Next company/interview]

After Advancing to Next Round

## Advancement Debrief

**Next round:** [Type of interview]
**Scheduled for:** [Date/time]

**What worked in this round (do more of):**
- [Strength 1]
- [Strength 2]

**What to prepare for next round:**
- [Prep item 1: e.g., "They mentioned wanting deeper metrics analysis"]
- [Prep item 2: e.g., "Next round is with engineering - brush up on technical"]

**Research to do before next round:**
- [ ] [Task 1]
- [ ] [Task 2]

After Take-Home Assignment Presentation

## Assignment Presentation Debrief

**Assignment type:** [Product spec / Data analysis / Design exercise / etc.]

**Time invested:** [Hours spent on assignment + presentation]

**Presentation format:** [Slides / Doc walkthrough / Live demo / etc.]

**Questions they asked:**
1. [Question 1 - and how I answered]
2. [Question 2 - and how I answered]

**What they liked (based on feedback/reactions):**
- [Positive 1]
- [Positive 2]

**What they pushed back on:**
- [Concern 1: e.g., "Questioned my prioritization approach"]
- [Concern 2]

**If I could redo the assignment:**
- [Change 1]
- [Change 2]

Pro Tips

  1. Write immediately: Don't wait 30 min or you'll forget half
  2. Be specific: "I rambled" → "Spent 12 min on intro instead of 5 min"
  3. Track patterns: After 5 interviews, review all debriefs for themes
  4. No sugar-coating: Brutal honesty helps more than fake positivity
  5. Celebrate wins: Note what worked so you can repeat it
  6. Quantify: "Felt rushed" → "Had 5 min left, needed 10 min"
  7. Ask for feedback: If recruiter offers, take detailed notes

Common Mistakes to Avoid

❌ Waiting hours/days to debrief (memory fades) ✅ Capture within 15 minutes while fresh

❌ Generic notes ("did okay on metrics") ✅ Specific observations ("used DAU/MAU ratio, interviewer nodded")

❌ Only noting negatives (demotivating) ✅ Balance strengths and weaknesses

❌ No action items (passive reflection) ✅ Concrete improvements for next time

❌ Ignoring patterns across interviews ✅ Review all debriefs monthly for themes


Questions to Ask Yourself

After every interview debrief:

  1. Preparation: Did I do enough company research?
  2. Structure: Did I use a framework or wing it?
  3. Specificity: Were my metrics and examples concrete?
  4. Time: Did I manage time well or rush/ramble?
  5. Confidence: Did I sound certain or hedge too much?
  6. Learning: What's the ONE thing I'll do differently next time?

Integration with Other Skills

Before interview:

  • /interview-prep - Preparation checklist

During job search:

  • Track all feedback in outputs/interview-feedback/
  • Monthly review: Look for patterns
  • Adjust prep based on recurring weaknesses

After getting offer:

  • Review all debriefs to see what worked
  • Document lessons for future job searches


Output Quality Self-Check

Before delivering the debrief, verify:

Check Criteria Pass?
Emotional check-in completed Asked how the PM is feeling before diving into scoring [ ]
5 dimensions scored All 5 core dimensions (Framework, Specificity, Creativity, Communication, Product Sense) have a 1-5 score [ ]
Scores are honest Scores reflect actual performance, not what the PM wants to hear [ ]
Specific examples Each strength and weakness cites a specific moment from the interview, not generalities [ ]
Improvement drills provided At least 1 targeted drill for the lowest-scoring dimension [ ]
Pattern check done Checked outputs/interview-feedback/ for prior debriefs and showed trends if found [ ]
Action items concrete Next steps are specific and time-bound, not vague ("practice more") [ ]
Follow-up tasks included Thank-you email reminder, LinkedIn connection, next round prep noted [ ]
Output file saved Debrief saved to outputs/interview-feedback/[date]-[company]-debrief.md [ ]

If any check fails, address it before delivering the output.


Remember: Every interview is data. Even rejections teach you something. Capture it all, learn fast, and improve between each round.


Sources:

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results