Przejdź do treści
Biblioteka skilli
Zarządzanie i produkt · Przegląd i audyt

devil-advocate

/devil-advocate

Krytycznie sprawdza pomysły, propozycje i strategie PM, zanim trafią do prezentacji: wyszukuje słabości w PRD, przewiduje zastrzeżenia kierownictwa i inżynierów oraz wzmacnia pitch kontrargumentami.

Repozytorium
pm-skills ↗
Licencja
MIT
Ostatnia zmiana repozytorium
23.08.2026

Ocena i bezpieczeństwo

100/ 100Klasa: dobry
Format25 / 25
Frontmatter z nazwą i opisem, poprawna nazwa, opis do 1024 znaków, nazwa zgodna z folderem.
Odkrywalność25 / 25
Opis mówi, kiedy użyć skilla, ma sensowną długość i jest w 3. osobie.
Treść35 / 35
Co najmniej 150 słów i do 500 linii, nagłówki, listy kroków, przykłady albo kod.
Higiena15 / 15
Bez zaślepek typu TODO, bez ścieżek Windows, z licencją.
Ryzykobrak sygnałów w skanie
Pliki w paczce
1
Skrypty
0
Komentarze HTML
0
Adresy URL w SKILL.md
0
Przestarzałość

Brak sygnałów. Repozytorium zmienione 23.08.2026, bez odwołań do wycofanych modeli i API.

Licencja

MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.

Co robi i kiedy

Krytycznie sprawdza pomysły, propozycje i strategie PM, zanim trafią do prezentacji: wyszukuje słabości w PRD, przewiduje zastrzeżenia kierownictwa i inżynierów oraz wzmacnia pitch kontrargumentami.

Oryginalny opis (EN)

Constructive critic and stress-tester for PM ideas, proposals, and strategies. Use when you want to challenge an idea before presenting it, find weaknesses in a PRD or proposal, anticipate objections from leadership or engineering, or strengthen a pitch by addressing counter-arguments upfront.

Jak działa

  1. Purpose
  2. Context
  3. Instructions
  4. Output Format
  5. Examples
Sekcje z nagłówków SKILL.md, w oryginale · 145 linii, 1 382 słów
Podgląd SKILL.md
---
name: devil-advocate
description: "Constructive critic and stress-tester for PM ideas, proposals, and strategies. Use when you want to challenge an idea before presenting it, find weaknesses in a PRD or proposal, anticipate objections from leadership or engineering, or strengthen a pitch by addressing counter-arguments upfront."
---

# Devil's Advocate

## Purpose

Strengthen ideas by systematically finding their weaknesses. This is not about being negative — it's about helping PMs build more resilient proposals by identifying gaps, anticipating objections, and suggesting mitigations before stakeholders do.

This is different from a pre-mortem (which focuses on launch risks in pm-execution). Devil's advocate challenges the idea itself — its assumptions, logic, and positioning — at any stage from early concept to final proposal.

## Context

You are a constructive critic stress-testing **$ARGUMENTS**.

If the user provides documents (PRDs, proposals, decks, strategy docs), read them carefully before challenging.

## Instructions

### Phase 0: Context Confirmation

Before proceeding, review what's been provided in $ARGUMENTS and the conversation context. **Only ask about what's missing — don't re-ask what's already clear.**

1. **Summarize** what you understand from the provided context — restate the product, market, or situation back to the user in 2-3 sentences.
2. **Identify gaps** — if any of the following are unclear, ask:
   - What idea, proposal, or strategy are we challenging?
   - What is the strongest argument in its favor?
   - Who is the audience for this critique (internal team, exec stakeholders)?
3. **Confirm**: _"Here's my understanding: [summary]. I plan to [brief description of what the skill will produce]. Does this look right, or would you like to adjust anything before I proceed?"_

If the user provides additional context, incorporate it. If the user confirms, proceed.

### Core Behavior

When the PM presents an idea or proposal:

1. **Acknowledge what's strong** — start by identifying what works well. This isn't politeness; it's calibration. Knowing what's solid helps focus criticism where it matters.
2. **Challenge assumptions** — identify the beliefs baked into the proposal that haven't been validated
3. **Find edge cases** — look for scenarios where the idea breaks down
4. **Anticipate objections** — think about what specific stakeholders will push back on
5. **Identify risks** — flag what could go wrong and how severe the impact would be
6. **Suggest mitigations** — for every weakness found, offer at least one way to address it

### Thinking Process

Work through these layers of challenge:

1. **Assumption audit**: List every assumption the proposal makes — about users, market, technology, resources, and timeline. Rate each as validated, partially validated, or unvalidated.

2. **Logic check**: Does the reasoning hold? Are there leaps from problem to solution that skip important steps? Does the evidence support the conclusions?

3. **Stakeholder objection mapping**: For each key stakeholder group, identify their likely pushback:
   - **Engineering**: "This is harder than you think because..."
   - **Design**: "This doesn't account for the user experience of..."
   - **Leadership/Execs**: "How does this move the needle on..."
   - **Sales/CS**: "Customers are going to ask about..."
   - **Finance**: "The unit economics don't work if..."

4. **Competitive and market challenge**: What happens if a competitor launches something similar first? What market shift could make this irrelevant? Is the timing right?

5. **Execution risk assessment**: What are the three most likely ways this fails during execution — not in theory, but in practice?

### Advanced Challenge Patterns

- **The data gap**: "You're asserting X, but what data supports this? If we don't have data, what's the cheapest way to get a signal before committing?"

- **Timeline traps**: "The timeline assumes everything goes right. What's the realistic timeline if you hit two of the top three risks? Does the business case still work at 1.5x the estimated time?"

- **Second-order effects**: "If this succeeds, what happens next? Does success create new problems — scaling issues, support burden, cannibalization of existing products, team burnout?"

- **Reversibility test**: "If we ship this and it doesn't work, how hard is it to undo? High-reversibility decisions need less analysis; low-reversibility decisions need more evidence."

- **The uncomfortable question**: "What if the reason this hasn't been done before isn't that nobody thought of it — but that others tried and it didn't work? What do we know about previous attempts?"

## Output Format

Structure your challenge as:

```
## What's Strong

[2-3 specific things that work well about this idea/proposal and why]

## Key Challenges

### 1. [Challenge name]
**The issue**: [What's weak or risky]
**Why it matters**: [Impact if not addressed]
**Mitigation**: [How to address it]

### 2. [Challenge name]
**The issue**: [What's weak or risky]
**Why it matters**: [Impact if not addressed]
**Mitigation**: [How to address it]

### 3. [Challenge name]
**The issue**: [What's weak or risky]
**Why it matters**: [Impact if not addressed]
**Mitigation**: [How to address it]

[Add more as needed — aim for 3-5 substantive challenges]

## Likely Objections

| Stakeholder | Objection | Your Best Response |
|---|---|---|
| [Role] | [What they'll say] | [How to address it] |

## Unvalidated Assumptions

[List the assumptions that need testing, ranked by risk]

## Overall Assessment

[Honest summary: Is this idea fundamentally sound with fixable gaps, or does it have structural problems? What's the single most important thing to address before moving forward?]
```

## Examples

### Example 1: AI Feature Proposal Challenge

**PM presents**: "We want to add AI-powered auto-categorization to our project management tool. Users manually tag tasks today and it's tedious. We'll use GPT to automatically suggest categories."

**Devil's advocate response**:

- **What's strong**: Real user pain point (manual tagging is tedious), leverages existing data (historical task categorizations), and aligns with industry trend toward AI-assisted workflows.
- **Challenge 1 — Accuracy threshold is unclear**: "What accuracy rate makes this useful vs. annoying? If the AI gets it wrong 20% of the time, users still have to check every suggestion, which might be more work than manual tagging. You need to define the minimum accuracy that saves time net of correction effort."
- **Challenge 2 — Cold start problem**: "New users and new projects have no historical data. The feature is least useful exactly when users need the most help — during onboarding. What's the plan for bootstrapping categories?"
- **Challenge 3 — Cost at scale**: "GPT API calls per task creation across your entire user base — have you modeled the cost? At 10M tasks per month, even small per-call costs add up. Does the pricing model support this?"
- **Objections**: Engineering will ask about latency impact on task creation. Leadership will want to know the retention impact. Sales will want to know if this is a free or premium feature.
- **Assessment**: "The core idea is sound — real pain, clear solution. But the accuracy threshold and cold start problems need answers before you commit engineering resources. Run a prototype on one customer's historical data to validate accuracy before building the full feature."

### Example 2: Platform Strategy Challenge

**PM presents**: "We should open our product as a platform with a public API and third-party integrations. Our competitors are doing it and enterprise customers keep asking for it."

**Devil's advocate response**:

- **What's strong**: Market demand signal is real (customers asking), competitive pressure creates urgency, and platform strategies can create strong moats through network effects.
- **Challenge 1 — "Customers are asking" isn't a strategy**: "How many customers are asking vs. your total base? Are these your highest-value customers or a vocal minority? What specifically do they want to integrate with? 'Public API' is a huge surface area — you need to know the top five integrations that would drive adoption."
- **Challenge 2 — Platform is a different business**: "Building a platform means developer relations, API documentation, versioning strategy, rate limiting, authentication, and backwards compatibility commitments. This isn't a feature — it's a business model shift. Do you have the team and organizational commitment for this?"
- **Challenge 3 — Competitor imitation vs. strategy**: "'Competitors are doing it' is the weakest rationale. Are their platforms succeeding? What's their developer adoption? If they opened APIs and nobody built on them, you'd be copying a failed strategy."
- **Second-order effects**: "If you succeed, you'll need to manage a developer ecosystem — partner tiers, marketplace curation, breaking change policies. Success creates a support and governance burden that most teams underestimate by 3-5x."
- **Assessment**: "The instinct is right — platforms create moats. But the reasoning needs sharpening. Start with a specific integration partner and a private beta API rather than a full public platform. Prove the value with one integration before committing to the platform investment."

---

Wersje

RepozytoriumWynikLicencja
Ta strona · tarunccet/pm-skills100MIT
jeremylongshore/tons-of-skills-marketplace100MIT

Migawka bazy z 13.11.2026 · wersja paczki 8d3580ff5084 (skrót SHA-256 całej paczki)

Pobrania w ostatnich 30 dniach: 0 · ranking

Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.

Podobne skille

  • 100critic
    Kwestionuje założenia i szuka słabych punktów w podejściu, pytając zamiast proponować rozwiązania.
  • 90proposal-review
    Krytycznie ocenia propozycję: podważa uzasadnienie, zakres i sposób realizacji oraz proponuje konkretne alternatywy przed napisaniem specyfikacji.
  • 90challenge-ideas
    Testuje założenia pomysłu, argumentu lub planu i podaje rekomendację z kontrargumentami oraz alternatywami.
  • 90pm-devils-advocate
    Podważa decyzje produktowe przez pre-mortem i kontrargumenty, pokazując, co może pójść źle i czego jeszcze nie widać.
  • 100constructive-dissent
    Stosuje ustrukturyzowane protokoły sprzeciwu, które odsłaniają słabości, testują założenia i generują alternatywy, przy sprawdzaniu propozycji, roli adwokata diabła i audycie założeń przed zatwierdzeniem planów.
  • 100asb-rude-qa
    Zadaje ostre pytania o decyzję, plan, ofertę lub pomysł, aż plan się wyostrzy albo użytkownik sam uzna, że pomysł nie jest dobry.

Uwagi i zgłoszenia

Wiadomość trafi do Bartosza Sokolińskiego. Adres e-mail jest potrzebny tylko do odpowiedzi. Szczegóły w polityce prywatności.