| name | prototype-gdd-generator |
| description | Generate structured Game Design Documents for prototypes through interactive Q&A. Use this when the user wants to document a game concept, create a prototype GDD, or formalize their design ideas. |
| domain | design |
| type | generator |
| version | 1.0.0 |
| allowed-tools | ["Write"] |
Prototype GDD Generator Skill
This skill generates comprehensive Game Design Documents for prototype games through an interactive questioning process, then outputs a structured markdown document following industry-standard prototype GDD format.
When to Use This Skill
Invoke this skill when the user:
- Asks to "create a GDD" or "write a design doc"
- Says "help me document my game idea"
- Wants to formalize a game concept into structured format
- Asks "how do I write a design document?"
- Says "generate a GDD for [game concept]"
- Needs to communicate game design to a team
- Wants to plan a prototype systematically
Core Principle
Structured documentation enables focused prototyping:
- ✅ Forces designer to articulate what they're testing
- ✅ Separates core mechanics from scope creep
- ✅ Defines success criteria before building
- ✅ Creates shared understanding for team
- ✅ Provides decision framework during development
- ✅ Documents design intent for future reference
Interactive Questioning Workflow
Phase 1: Core Concept (Required)
Always start with these questions:
I'll help you create a Prototype Game Design Document! Let's start with the core concept:
**1. Game Concept**
In 1-2 sentences, what is your game?
(Example: "A tactical auto-battler where you program weapon AI before battle, then survive waves while controlling movement")
**2. Elevator Pitch**
If you had 30 seconds to pitch this, what would you say?
(Combine genre, core mechanic, and hook)
**3. Design Pillars**
What are the 2-4 core principles that define this game?
(Example: "Tactical Programming Over Reflexes", "Emergent Complexity from Simple Rules")
**4. Primary Influences**
Which 2-5 games inspire this concept?
For each, explain WHAT you're taking from it.
(Example: "Vampire Survivors - 15-min runs with level-up choices, but NOT the direct aiming")
Wait for user response before proceeding.
Phase 2: Prototype Questions (Critical Testing Focus)
After receiving Phase 1 answers, ask:
Great! Now let's define what this prototype needs to TEST:
**5. Critical Questions**
What are the 3-5 core questions this prototype must answer?
Format: "Does [mechanic] feel [quality]?"
Examples:
- "Does pre-battle programming feel strategic (not arbitrary)?"
- "Is the execution phase satisfying to watch?"
- "Can players understand why they won/lost?"
**6. Success vs Failure**
For each critical question, what does SUCCESS look like? What does FAILURE look like?
Example:
Q: "Does programming feel strategic?"
Success = Players discuss rule choices, experiment with setups
Failure = Players ignore rules, pick randomly
**7. Timeline**
How long do you have for this prototype?
- [ ] Weekend (2-3 days)
- [ ] Week (5-7 days)
- [ ] Two weeks
- [ ] Month
- [ ] Other: _______
Wait for user response before proceeding.
Phase 3: Core Mechanics (Deep Dive)
After receiving Phase 2 answers, ask:
Perfect! Now let's break down the mechanics:
**8. Core Mechanic #1**
What is the PRIMARY mechanic players interact with?
(Example: "Weapon Rule System - assign targeting rules to weapons")
Describe:
- How it works (step-by-step)
- Player inputs required
- System outputs/feedback
- Why it's interesting
**9. Supporting Mechanics**
What are the 2-4 supporting mechanics?
(Example: "WASD movement", "Wave-based enemy spawning", "Level-up system")
For each:
- Brief description
- How it interacts with core mechanic
**10. Content Scope**
How many of each content type?
- Player abilities/weapons/actions: X
- Enemy types: X
- Levels/arenas: X
- Win condition: (describe)
- Lose condition: (describe)
Wait for user response before proceeding.
Phase 4: Scope Definition (What's In/Out)
After receiving Phase 3 answers, ask:
Almost there! Let's define the scope boundaries:
**11. What's IN the Prototype** (Minimum Viable)
What MUST be in the prototype to test your critical questions?
List features/systems required for meaningful playtest.
**12. What's OUT of the Prototype** (Explicitly Excluded)
What are you intentionally NOT building?
These are features that would be nice but distract from testing core questions.
For each OUT item, explain WHY it's excluded.
(Example: "❌ Multiple arenas - Reason: Testing gameplay, not variety")
**13. Target Playtime**
How long should a single playthrough take?
(Example: "15 minutes per run" or "5-minute rounds")
Wait for user response before proceeding.
Phase 5: Implementation & Success Metrics
After receiving Phase 4 answers, ask:
Final questions!
**14. Implementation Phases**
Break your timeline into phases. For each phase:
- Goal (what you're building)
- Deliverables (specific features)
- Test criteria (how you know it works)
Example:
Phase 1 (Day 1 Morning - 4h): Movement & Basic Combat
- Goal: Player moves, weapon auto-fires
- Deliverables: WASD movement, one weapon, enemy spawning
- Test: Can kill enemies with basic weapon
**15. Success Metrics**
How will you MEASURE if the prototype succeeded?
Qualitative (playtester observations):
- What should playtesters SAY? (quotes)
- What should playtesters DO? (behaviors)
Quantitative (numbers):
- Session length target
- Completion rate target
- "One more run" rate
**16. Known Risks**
What could go wrong? For each risk:
- Risk: [What might fail]
- Mitigation: [How to prevent it]
- Fallback: [What to do if it happens anyway]
Example:
Risk: Rules feel arbitrary, don't affect outcome
Mitigation: Make enemies explicitly weak to certain rules
Fallback: Add "Effective" tag when rule matches enemy
Wait for user response before proceeding.
Phase 6: Generate GDD
After gathering all answers, generate the full GDD.
Use this structure:
# [GAME TITLE] - Prototype Design Document
**Version:** [Version] - [Prototype Type]
**Goal:** [One-sentence goal from critical questions]
**Timeline:** [Timeline from answers]
**Date:** [Current date]
---
## 1. CONCEPT
### Elevator Pitch
[From Phase 1, Q2]
### Design Pillars
[From Phase 1, Q3 - expand each pillar with 2-3 sentences]
**[Pillar 1 Name]**
[Description and why it matters]
**[Pillar 2 Name]**
[Description and why it matters]
### Primary Influences
[From Phase 1, Q4]
**[Game 1]**
- What we're taking: [Specific mechanic/feel]
- Applies: [How it applies to our game]
**[Game 2]**
[Same format]
---
## 2. WHAT WE'RE TESTING
### Critical Questions
[From Phase 2, Q5 - numbered list]
**Q1: [Question]**
- Success = [Success criteria]
- Failure = [Failure criteria]
**Q2: [Question]**
[Same format]
### Success Criteria
[From Phase 5, Q15]
After [timeline] playtest, score each question 1-5:
- 1 = Doesn't work at all
- 3 = Works but needs improvement
- 5 = Works great, build full game
**Decision Threshold:**
- [X-Y] points: Build full game
- [X-Y] points: Iterate and retest
- <[X] points: Pivot or kill
---
## 3. CORE MECHANICS
### [Core Mechanic Name]
[From Phase 3, Q8 - detailed description]
**How It Works:**
1. [Step 1]
2. [Step 2]
3. [Step 3]
**Player Inputs:**
[What player does]
**System Response:**
[What game does in response]
**Interactions:**
[How this connects to other mechanics]
### [Supporting Mechanic 1]
[From Phase 3, Q9]
**Specifics:**
[Detailed breakdown]
**Interactions:**
[How it connects to core mechanic]
### [Supporting Mechanic 2]
[Same format for each supporting mechanic]
---
## 4. CONTENT SCOPE
### [Content Type 1] ([Number] types)
[From Phase 3, Q10]
**[Item 1]:**
- [Stats/attributes]
- [Behavior]
- [Purpose/role]
**[Item 2]:**
[Same format]
### [Content Type 2]
[Same format for enemies/levels/etc.]
---
## 5. PROTOTYPE SCOPE
### What's IN (Minimum Viable)
[From Phase 4, Q11]
**[System Category 1]:**
- [Feature 1]
- [Feature 2]
**[System Category 2]:**
[Same format]
### What's OUT (Not for Prototype)
[From Phase 4, Q12]
**❌ [Feature Name]**
- Reason: [Why excluded]
- Prototype: [What you're doing instead]
**❌ [Feature Name]**
[Same format]
---
## 6. IMPLEMENTATION PHASES
### [Phase 1 Name] ([Timeline] - [Hours])
[From Phase 5, Q14]
**Goal:** [What you're building]
**Deliverables:**
- [Feature 1]
- [Feature 2]
- [Feature 3]
**Test:** [How you know it works]
---
### [Phase 2 Name]
[Same format for each phase]
---
## 7. SUCCESS METRICS
### Playtester Observations ([When])
[From Phase 5, Q15 - qualitative]
**During [Phase/Screen]:**
- Do they [behavior]? (Good: [example] Bad: [example])
- Do they [behavior]? (Good: [example] Bad: [example])
**After Run:**
- Ask: "[Question]" (Want: [desired answer])
- Ask: "[Question]" (Want: [desired answer])
### Quantitative Targets
[From Phase 5, Q15 - quantitative]
- [Metric]: [Target] ([context])
- [Metric]: [Target] ([context])
---
## 8. RISK MITIGATION
[From Phase 5, Q16]
**Risk: [Risk description]**
- Mitigation: [How to prevent]
- Fallback: [What to do if happens]
**Risk: [Risk description]**
[Same format]
---
## 9. POST-PROTOTYPE DECISION TREE
### If Score [High Range] (Build Full Game)
**Immediate Next Steps:**
- [Feature to add]
- [System to expand]
- [Content to create]
**Timeline:** [Estimated timeline to v1.0]
### If Score [Medium Range] (Iterate)
**Identified Issues → Solutions:**
- [Issue] → [Solution]
- [Issue] → [Solution]
**Timeline:** [Iteration timeline], retest, reassess
### If Score [Low Range] (Pivot/Kill)
**Exit Criteria:**
- [Fundamental flaw 1]
- [Fundamental flaw 2]
**Pivot Options:**
- [Alternative approach 1]
- [Alternative approach 2]
---
**END OF DESIGN DOCUMENT**
This prototype tests: [Restate core question from Section 2]. By [end date], you'll know.
Output File Naming
Save GDD to:
docs/[game-name-kebab-case]-prototype-gdd.md
Examples:
docs/mech-survivors-prototype-gdd.md
docs/tower-defense-prototype-gdd.md
docs/deck-builder-prototype-gdd.md
Section Guidelines
Concept Section
- Elevator Pitch: 1-2 sentences max, focus on hook
- Design Pillars: 2-4 pillars, each gets 3-5 sentence explanation
- Influences: Be specific about WHAT you're taking (not just "inspired by X")
What We're Testing Section
- Critical Questions: Format as "Does X feel Y?" or "Can players do Z?"
- Success Criteria: Concrete observable behaviors, not vague feelings
- Decision Threshold: Numerical scoring system (forces honesty)
Core Mechanics Section
- One primary mechanic: The thing the prototype tests
- 2-4 supporting mechanics: What makes primary mechanic work
- Detailed breakdown: How it works step-by-step, not just description
Prototype Scope Section
- What's IN: Only what's needed to answer critical questions
- What's OUT: Explicitly name excluded features to prevent scope creep
- Justify exclusions: Every OUT item needs a reason
Implementation Phases Section
- Break timeline into phases: 2-6 phases depending on total timeline
- Each phase has goal + deliverables + test: Concrete milestones
- Time estimates: Hour estimates for accountability
Success Metrics Section
- Qualitative observations: What to watch for during playtests
- Quantitative targets: Numbers to hit (completion rate, session length)
- Post-playtest questions: Specific questions to ask playtesters
Risk Mitigation Section
- Identify risks early: What could make prototype fail to answer questions
- Mitigation first: How to prevent the risk
- Fallback second: What to do if mitigation doesn't work
Decision Tree Section
- Three outcomes: Build full game, iterate, pivot/kill
- Concrete score ranges: Based on success criteria scoring
- Next steps for each: What to do in each scenario
Important Guidelines
Keep It Prototype-Focused
- ❌ Don't plan full game features
- ✅ Focus on testing core questions
- ❌ Don't include monetization, meta-progression, etc.
- ✅ Include only what's needed for meaningful playtest
Be Ruthlessly Specific
- ❌ Vague: "Good controls"
- ✅ Specific: "WASD movement at 200 px/s"
- ❌ Vague: "Fun weapons"
- ✅ Specific: "5 weapons: Machine Gun (5 dmg, 5/s, 300px range)..."
Define Success Measurably
- ❌ "Players like it"
- ✅ "Players play 3+ runs in first session"
- ❌ "Combat feels good"
- ✅ "Players can explain why they won/lost after run"
Scope Ruthlessly
- Default to LESS for prototype
- Every feature should answer a critical question
- If it doesn't test core loop, it's OUT
- You can add it to the full game later
Example Invocations
User: "Help me write a GDD for my tower defense game"
User: "Create a design document for my roguelike prototype"
User: "I need to document my game idea"
User: "Generate a GDD based on this concept: [describes game]"
User: "How do I write a design doc for a weekend prototype?"
Workflow Summary
- Ask Phase 1 questions (concept, pitch, pillars, influences)
- Wait for user response
- Ask Phase 2 questions (critical questions, success criteria, timeline)
- Wait for user response
- Ask Phase 3 questions (core mechanics, supporting mechanics, content)
- Wait for user response
- Ask Phase 4 questions (scope: what's in/out, target playtime)
- Wait for user response
- Ask Phase 5 questions (implementation phases, metrics, risks)
- Wait for user response
- Generate complete GDD following template structure
- Save to
docs/[game-name]-prototype-gdd.md
- Confirm file location to user
Tips for Quality GDDs
For the User
- Be honest about timeline: Don't plan 2 weeks of work for a weekend
- Answer "why": Every mechanic should have a purpose
- Think about testing: How will you know if it works?
- Cut ruthlessly: Prototype scope should feel uncomfortably small
For Claude
- Ask follow-up questions: If answer is vague, ask for specifics
- Push back on scope creep: If user includes too much, ask which features test critical questions
- Suggest improvements: If critical questions are vague, help sharpen them
- Use examples: Reference similar games to clarify unclear concepts
Quality Checklist
Before finalizing GDD, verify:
- ✅ Every mechanic maps to a critical question
- ✅ Success criteria are observable behaviors (not feelings)
- ✅ Timeline is realistic (not aspirational)
- ✅ Scope has clear IN/OUT boundaries with justifications
- ✅ Implementation phases have concrete deliverables
- ✅ Risks have both mitigation AND fallback plans
- ✅ Decision tree has numerical thresholds
- ✅ All sections use specific numbers/details (not vague descriptions)
This skill transforms vague game ideas into structured, testable prototype plans. The interactive workflow ensures the designer thinks critically about what they're building and why.